Skip to main content

Full text of "Agenda setting and political discourse : the framing of political campaign reform in 1999"

See other formats


AGENDA SETTING AND POLITICAL DISCOURSE: 
THE FRAMING OF POLITICAL CAMPAIGN REFORM IN 1999 






BY 
SHANNON L. BICHARD 



A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE 

UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE 

REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY 

UNIVERSITY OF FLORIDA 



2001 



ACKNOWLEDGEMENTS 

First and foremost, I wish to thank the Lord Almighty for bestowing upon 
me the talent and most importantly the tenacity to reach all of my academic 
goals. This dissertation was produced while enduring some of the most 
difficult circumstances, yet He gave me the strength to continue even in my 
times of weakness. 

My sincere gratitude also extends to my family and friends for enduring my 
frequent irritable moods while completing this project. To my mother, I 
profess the ultimate "thank you" for being my mentor and intellectual stimulant 
for as long as I can remember. I will have succeeded in life if I can one day 
become even half the woman that she is. To my husband Rob, I owe so 
much. He has been literally by my side with support and encouraging words 
convincing me again and again that I could succeed if I would just take "one 
step at a time." 

Finally, my gratefulness extends to my advisor, Dr. Marilyn Roberts, for 
her impeccable advice and guidance. She has set for me an example that I 
wish to emulate as I interact with future students and colleagues. To the rest 
of my committee, Dr. Martinez, Dr. Sutherland, Dr. Tipton, and Dr. Wanta, I 
am thankful for their patience in working with my schedule and long distance 
commute. Their support and critique have been invaluable. 



TABLE OF CONTENTS 

page 

ACKNOWLEDGEMENTS ii 

ABSTRACT v 

CHAPTERS 

1 INTRODUCTION 1 

Overview 1 

Research Purpose 3 

Dissertation Outline 4 

2 LITERATURE REVIEW 6 

Campaign Reform 6 

Public Opinion 8 

Agenda Setting 9 

The Public Agenda 10 

The Media Agenda 19 

The Policy Agenda 26 

The Issue Attention Cycle 28 

Issue Dynamics and Competition 30 

Framing 34 

Research Questions 51 

3 METHODOLOGY 62 

The Issue Attention Cycle 63 

The Media Agenda 67 

The Internet 70 

The Public Agenda 74 

The Congressional Agenda 77 

Coding Categories 79 



4 RESULTS 83 

Descriptive Results of Content Analysis 83 

Research Questions 90 

5 DISCUSSION 118 

APPENDICES 

A SAMPLE ARTICLE FROM THE NEW YORK TIMES 130 

B SAMPLE "MOST IMPORTANT PROBLEM" POLLS FOR 

THE CAMPAIGN REFORM ISSUE 132 

C SAMPLE MESSAGE FROM DEJA.COM SEARCH 

ENGINE 137 

D SAMPLE SPEECH FROM CONGRESSIONAL 

RECORD 139 

E CODING SHEETS FOR FRAMING CATEGORIES 141 

REFERENCES 151 

BIOGRAPHICAL SKETCH 169 



IV 



Abstract of Dissertation Presented to the Graduate School 
of the University of Florida in Partial Fulfillment of the 
Requirements for the Degree of Doctor of Philosophy 

AGENDA SETTING AND POLITICAL DISCOURSE: 
THE FRAMING OF POLITICAL CAMPAIGN REFORM IN 1999 

By 

Shannon L. Bichard 

August 2001 

Chair: Marilyn S. Roberts 

Major Department: Mass Communication 

Political campaign reform is an issue of great interest to many Americans. 
This dissertation examines the shaping of the political climate for the issue by 
applying first and second level agenda setting analysis. Content analysis 
was used as a method to procure the agendas of the media, public, and 
Congress. Articles from the New York Times and the Washington Post were 
used to depict the media agenda. Speeches on the floor of the House and 
Senate portrayed the congressional agenda. Finally, messages from 
USENET on-line discussion groups represented the public agenda for the 
issue. 

Data were accumulated regarding campaign reform for the 52 weeks of 
1999. The frequency of emphasis in each agenda revealed first level effects 
in order to establish the direction of influence regarding the salience of the 



issue. This examination was then expanded to include an analysis of the 
attributes used by each agenda to discuss the issue. Attributes were coded 
using a multidimensional framing approach. The dimensions included the 
sub-issues, affective, and cognitive elements used to frame the issue. The 
overall objective of the investigation was to capture a complete and accurate 
picture of the forces that sculpted the three agendas. 

Autoregressive integrated moving averages (ARIMA) time series analysis 
was utilized as a statistical tool for modeling each univariate series of data. 
Cross-correlation analysis was then performed to examine the relationships 
among the agendas under scrutiny. A variety of significant effects were 
noted, but some underlying trends did emerge. Results suggest that the 
public have a powerful influence on the salience attributed to the issue of 
campaign reform by the Congress and the media. This influence persists with 
respect to the public agenda leading the media in their choice of attributes. 
The framing attributes used by the Congress are also notably influential in 
subsequent attributes used in media coverage for the issue. These findings 
both align and depart from past agenda setting research, but most assuredly 
shed light on the need for the future development of a multi-level approach to 
the study of agenda setting. 



VI 



CHAPTER 1 
INTRODUCTION 



Overview 

Freedom of Speech. Political Corruption. Incumbent Advantage. Soft 
Money. Issue Advocacy. Contribution Limits. These are all familiar terms in 
the on-going campaign reform debate. Discussion of such reform has 
permeated the mass media. With the increased attention afforded the issue 
in the 2000 election, campaign reform will likely persist in capturing the 
attention of journalists and Americans in general. 

As with most public issues, campaign reform has many facets. Because 
of its controversial nature, the issue also maintains a wide variety of 
proponents. To understand the meaning of campaign reform and make an 
accurate judgment regarding its applicability in the political process, the 
general public must be adequately informed on the issue. 

Often times, as Lippmann (1922) would suggest, when an individual 
desires information that is out of his realm of expertise or personal 
experience, he/she looks to media reports to build "a trustworthy picture 
inside his head of the world beyond his reach" (p. 1 8). This is especially true 
for political issues (Lazarsfeld, Berelson, & Gaudet, 1944; McCombs & Shaw, 



1972; McCombs & Weaver, 1973; Weaver, McCombs & Spellman, 1975; 
Weaver & McCombs, 1978). 

According to agenda setting literature, the salience attributed to issues in 
the media has a profound affect on those that audience members deem 
important (McCombs & Shaw, 1972; Behr & Iyengar, 1985; Shaw& 
McCombs, 1989; Rogers, Dearing & Chang, 1991). The media, by 
emphasizing certain issues, tell people what to think about (Cohen, 1963). 

Ideally, the mass media should provide its citizenry with a complete and 
accurate depiction of social issues that affect their everyday lives (Gamson, 
Croteau, Hoynes, & Sasson, 1992). Political consciousness should be 
stimulated with a wide variety of views and alternative voices. Although this is 
the ideal, most would agree that the American media often falls short. Media- 
generated images encourage the general public to construct meaning in a 
way that, as Gamson suggests (1992, p.374), "evinces the power and point of 
view of the political and economic elites who operate and focus it." 

The organization of such information by the media and subsequent effects 
have astounded researchers for years (Lazarsfeld, Berelson, & Gaudet, 1944; 
Klapper, 1960; McCombs & Shaw, 1972; Rogers & Dearing, 1987; McCombs 
& Evatt, 1995). The specific manner in which the construction of 
communication texts can influence individual cognition by selective focus is 
often referred to as framing and for many is considered the second level of 
agenda setting (McCombs & Evatt, 1995; Ghanem, 1996). 



The theoretical premise for framing analysis can be found in both 
sociological and psychological research. Some view framing as the 
sociological process of constructing news, while others understand framing as 
the psychological dynamics of audience consumption (Goffman, 1974; Gitlin, 
1980; Tversky & Kahneman, 1982; Gamson & Lasch, 1983; Entman, 1991). 
Framing analysis moves beyond traditional agenda setting research in an 
effort to investigate the influence of media attributes on individuals. This 
second level of agenda setting provides a more detailed view of the "pictures 
in our heads" so eloquently mentioned by Lippmann (1922). 

Research Purpose 

It is the purpose of this study to investigate the effects of traditional 
agenda setting as well as second level agenda setting. The issue of political 
campaign reform will be analyzed. This issue has recently captured public 
and media attention and is sure to provide ample amounts of information for 
analysis. Specifically, data will be accumulated for the public, media, and 
congressional agendas. 

The current study will utilize a variety of traditional as well as new methods 
to operationalize each agenda. The media agenda will be represented by 
news coverage in the New York Times and the Washington Post. The public 
agenda will be represented by a pioneering analysis of Internet 
communication. The on-line discussion in USENET groups will be examined 
in order to provide a depiction of public opinion. The congressional agenda 



will be represented by congressional speeches on the floor of the United 
States House and Senate. Content analysis will serve as an unobtrusive tool 
with which to investigate the agenda setting effects for each of these 
agendas. 

The goal of this research is to better understand the transfer of salience 
with respect to the public, media, and congressional agendas. This includes 
the transfer of salience regarding which issues to think about as well as the 
transfer of salience regarding how to think about such issues. The campaign 
reform issue will be utilized in this analysis because of the heightened 
attention to the issue in recent months. The overall investigation is sure to 
provide a comprehensive assessment of the dynamic forces shaping the 
issue of political campaign reform. 

Dissertation Outline 

The following chapter will review pertinent literature in an effort to lay a 
foundation for the present research study. It will begin with an introduction to 
the basics of the campaign reform issue. This will serve as a starting point 
from which to begin a discussion of public opinion and the various sources 
that attempt to shape it. The theoretical background for agenda setting will 
then be provided. Specifically, the details involving the formation of the 
public, media, and policy agendas will be discussed. Next, the notion of issue 
attention and competition will be examined in an effort to highlight the 
importance of timing in agenda setting research. The second level of agenda 



setting will then be introduced with its roots in framing analysis and 
references to psychological and sociological literature. The chapter will 
conclude with research questions detailing the expected relationships at both 
the first and second levels of agenda setting. 

The third chapter will describe the methods chosen for the analysis. A 
discussion of each research question will be accompanied by a description 
and rationale for the chosen representation of each variable. The time frame 
for analysis will be discussed as well as the chosen method for examining the 
data. 

The fourth chapter will present the findings of the research study. This will 
be followed by a final chapter that summarizes and details the implications of 
the findings for future study. 






CHAPTER 2 
LITERATURE REVIEW 



Campaign Reform 

The American political system is firmly rooted in freedom and equality. 
The notion of "one person, one vote" is a valued American ideal. Since the 
appearance of political communities, money and politics have been firmly 
linked. While political equality is valued, the fact remains that money is not 
distributed equally. Achieving political equality in today's society relies a 
great deal on financial backing. Such dependence on money opens the door 
for financial power struggles and heated competition. Huge contributions and 
campaign expenditures along with thoughts of obligatory favoritism and 
abuses of power have led to an increased sense of worry in the minds of 
many Americans (Alexander, 1984). This, coupled with a decline in social 
capital and civic engagement, has led to a growing cynicism with American 
political figures and their campaigning behavior (Putnam, 1995). 

President Kennedy was the first modern president to consider the 
problematic tactics used in political campaigning. In 1961, he appointed a 
bipartisan Commission on Campaign Costs to investigate existing behavior 
and legislation (Alexander, 1984). At this point, the reform era had begun, but 
did not climax until 1971 with the passage of the Federal Election Campaign 



Act (FECA). This was the first comprehensive revision of federal campaign 
legislation since the Corrupt Practices Act of 1925. The act established a 
major turning point in the history of campaign finance reform. Among some of 
its provisions were limits on campaign media expenditures, limits on total 
candidate expenditures, contribution ceilings, and the requirement of 
disclosure reports detailing candidate funding, contributions, and 
expenditures. Amendments to the initial legislation were made in 1974, 1976, 
and 1979 to ensure that some segments of the legislation were not seen as 
inhibiting free speech or citizen participation. For example, in Buckley vs. 
Valeo (1976), the Supreme Court struck down the expenditure limitations, 
except for public financing in presidential elections, but upheld the 
contribution limits. Money spent is considered essential to political speech. 
The court established that spending money is speech and for this reason 
strict scrutiny must be given to any regulation of money contributed or spent 
in the political arena (Slabach, 1998). 

Recent efforts have been made to overhaul the existing reform legislation 
with the addition of stricter, more comprehensive regulations. Reform 
advocates claim they are responding to a public cynicism that is permeating 
our democracy (McCain, 1999). They claim that money is not speech, but a 
medium of exchange which facilitates speech (Slabach, 1998). Campaign 
reform proponents have issued bills that call for campaign issue advertising 



8 



regulation, a ban on soft money, full disclosure of financial information, and 
voluntary personal spending limits (Common Cause, 1998). 

Opposing activists maintain that campaign finance laws are a threat to the 
first amendment (Smith, 1997). They maintain that contributions are 
expressions of support for a candidate (Slabach, 1998). Such provisions 
stifle competition and perpetuate the status quo. Some argue that campaign 
finances are justifiably important to political campaigns, especially for non- 
incumbents (Jacobson, 1980). 

The controversy over campaign reform is yet to be settled. One thing is 
certain: intense debate regarding the issue is sure to continue as Americans 
assert their views in the aftermath of the 2000 election dispute. 

Public Opinion 

Public opinion is a powerful force in today's society. To harness this 
power, one must be able to define it. In the era before 1937 one would have 
been inclined to define public opinion as the views on public affairs held by an 
integration of community notables (Converse, 1987). This view suggests that 
public opinion "transcends individual opinion and reflects an abstract, 
common good" (Price, 1992, p.1 1). With the onset of more efficient means of 
mass communication and increased literacy, today the perceived definition of 
public opinion has been able to reach new dimensions. With increased 
technology, researchers can tally individual opinions on issues that include all 
citizens with an equal measure. Thus public opinion is more ideally reflective 



of a complete society. Although these measures are generally accepted in 
the research world, the precise nature of public opinion is still in many ways a 
mystery (Funkhouser, 1973). With many theories and suggestions given, no 
exclusively accepted definition of the term "public opinion" exists. This is not 
to say that public opinion is in any way meaningless. Its increased use in 
research of all kinds proves otherwise. We then must respect its many 
definitions with regard to different uses. Noelle-Neumann (1989, p.4) offers 
that "Public opinion is based on the unconscious striving of people living in a 
social unit to arrive at a common view, at the kind of agreement which is 
required to act and, if necessary, to make decisions." This denotation 
provides a comprehensive and concise definition of the term and will be used 
in the present study. In order to understand the complete process of public 
opinion formation with respect to issues, it is useful to analyze the agenda 
setting process. 

Agenda Setting 

Agenda setting research investigates the transfer of issue salience from 
one agenda to another. An issue can be conceptualized as something in 
dispute that offers multiple points of view (Becker, 1991 ). The agendas most 
often analyzed include the public agenda, the media agenda, and the policy 
agenda. Much research in agenda setting has attempted to include the 
interplay of all three factors (Gonzenbach, 1992; Jablonski & Bichard, 1997). 
With reference to these three agendas, Kosicki (1993, p. 101 ) notes that 



10 



"each part of the process is incomplete and somewhat unsatisfying by itself, 
but that by combining all three perspectives, the field can come closer to what 
a solid contemporary model of media influence ought to be." Many scholars 
agree that such an effort will help students and researchers to more clearly 
appreciate the connection between journalists, sources, public, and policy 
(McCombs, 1981, 1992; McCombs & Gilbert, 1986; Protess & McCombs, 
1991). Integrating the relationship among all three agendas provides a more 
complete view of the agenda setting process. Agenda setting historically 
originates with the investigation of the public agenda. 

The Public Agenda 

Much of the world remains out of reach for the average person. Therefore, 
when an individual desires information that is out of his/her realm of 
experience, he/she looks to the media to "make for himself a trustworthy 
picture inside his head of the world beyond his reach" (Lippmann, 1922, 
p. 18). For this reason, many issues which concern the public are in a large 
part determined by topics that are served to us by the media (Mackuen, 
1981). Funkhouser (1973, p.74) likens our attention to the media to playing a 
"crooked roulette wheel because it's the only wheel in town." The media- 
world experience must suffice because it is our only way of knowing what is 
happening in the world outside our immediate observation (Zucker, 1978). 
Without the use of the media by policy officials the government would have no 
way to communicate with the public. Information is vital for public and private 



11 



life, therefore, the modern world would be frightening without the media 
(Graber, 1980). Lippmann suggests that as events occur, the media focus on 
them and thus affect the formation of the public agenda. 

Early mass communication scholars began to study the effects of the 
media on public opinion change with little significant results. As they looked 
for the persuasive effects and came up short, the period became known as 
the "limited effects" era (Klapper, 1960). Many became discouraged and left 
the field (Beniger, 1987). Just when it seemed that the study of media effects 
was futile, communication researchers made a rediscovery. They began to 
approach media effects from a new angle. 

New research efforts indicated that the media may not be successful in 
changing opinions or "telling people what to think, but it is stunningly 
successful in telling its readers what to think about" (Cohen, 1963, p. 13). This 
twist focuses attention away from persuasion onto something clearly different. 
It suggests that although opinions are not formed or changed by the mass 
media, public agendas are set by the mass media. This move toward a new 
cognitive factor has intrigued researchers ever since. 

What is an agenda? Rogers and Dearing (1987, p.556) define an agenda 
as "a list of issues and events that are viewed at a point in time as ranked in a 
hierarchy of importance." The first researchers to use the phrase "agenda- 
setting" were McCombs and Shaw (1972). They studied the effects of the 
mass media on the public's political perceptions. Evidence showed that 



12 



voters tend to share common ideas on what is politically important with those 
stressed in the media. 

The Media's Influence 

The discussion of the media's political influence is not a new one. Several 
researchers in the "limited effects" era drew similar conclusions. A study by 
Lazarsfeld, Berelson, and Gaudet (1944) found a mobilizing effect of the 
media during a political campaign. However, it appears that interpersonal 
communication had more influence than the media on opinion formation. This 
indicates a "two-step" flow of communication from opinion leaders to the 
general public regarding the political issues. Opinion leaders appear to have 
more exposure to mass media, which they then pass along to less widely 
exposed publics (Katz & Lazarsfeld, 1955). Interpersonal communication and 
social interaction therefore play a large role in the public agenda setting 
process (Mcleod, Becker, & Byrnes, 1974; Wanta & Wu, 1992; Zhu, Watt, 
Snyder, Yan, & Jiang, 1993). In 1983 Noelle-Neumann developed a Strength 
of Personality (SP) scale to identify such influential opinion leaders. Brosius 
and Weimann (1996) tested the theory and found that those who ranked high 
on the scale were more likely to diffuse media information to the general 
public. This suggests that the media have the power to reach even those who 
do not watch, read, or listen. As Klapper (1960, p.30) states, "The common 
social habit of telling friends about mass communications which they may 



13 



themselves have missed appears likely to supplement the reinforcing 
capabilities of the original communications." 

The Influence of Individual Characteristics 

Several personality and lifestyle factors may also affect the reasons why 
some are more attuned or interested in the stories covered by the media. 
Agenda-setting effects have proven greater among individuals who were 
highly interested in an election, yet still uncertain in their choice of a candidate 
(McCombs & Weaver, 1973; Weaver, McCombs & Spellman, 1975; Mackuen 
& Coombs, 1981). This suggests that the media may serve as a point of 
orientation for individuals who feel disconnected with current issues. This 
function of the media promotes a sense of participation for consumers. In a 
time when the traditional roles of families are changing, some may turn to the 
mass media as an impersonal social tie to the community (Ball-Rokeach & 
Defleur, 1976). These findings agree with those of Blumler (1969), who 
proposed that audience orientations and motives influence and predict their 
media attentiveness. He suggests that different audience members are 
oriented to the mass media in different ways because of their unique 
personality dispositions and social circumstances. Blumler (1969) suggests 
that these orientations fall into three basic categories: cognitive (surveillance, 
vote guidance seeking, and reality exploration), diversion (entertainment, 
excitement), and personal identity needs (reinforcement, salience of personal 
experiences). Similarly, McGuire (1974, p. 182) argues that "mass 



14 



communications offer a wide range of gratifications in the form or content that 
is perceived as instructive regarding how to live, how to manage, what is 
happening, and what it means." The motivations of the individual may 
provide insight into the probability of media induced agenda setting effects. 
These effects may be contingent on an individual's need both for orientation 
and gratification (Chaffee, 1973). 

Education level may also have an effect on the agenda setting process. 
There is probably no single variable as influential with respect to agenda 
setting effects as education level (Converse, 1972). It may be suggested that 
the educated citizen is more sophisticated and critical. Well-educated 
individuals appear to show more selectivity in their media choices (Park & 
Kosicki, 1995). Perhaps then, their views are less influenced by the agenda 
setting efforts of the media (Iyengar & Kinder, 1987; Hill, 1985). This 
acknowledgment that all individuals are uniquely affected according to 
subjective experience suggests that the public must be disaggregated by 
attributes as much as possible in research situations (Rogers & Dearing, 
1987). 

In addition to uniquely affecting individuals in the public sector with respect 
to certain issues, the media may also influence an individual's inclination to 
speak up or keep quiet about issues. The spiral of silence suggests that 
individuals desire social approval and will not speak out if their views differ 
from that of the masses. Noelle-Neumann proposes that "To the individual, 



15 



not isolating himself is more important than his own judgment" (1974, p.43). 
Perhaps it is a widely accepted belief that the media agenda reflects the 
views of the majority. Therefore, those with a strong desire to conform will 
adhere to the saliency of the issues served to them by the media. This would 
then inhibit interpersonal flow of alternate media issues because of the fear of 
isolation. 

The Influence of Issue Characteristics 

Different issues in themselves may be characterized by their own unique 
agenda setting dynamics. Some issues are more easily conceptualized. 
Concrete issues require less cognitive development and possess less inertia, 
thus increasing the probability of agenda setting effects. On the other hand, 
ambiguous issues are unpredictable, difficult to visualize, and thus 
presumably less susceptible to the power of agenda setting (Yagade & 
Dozier, 1990; Protess, Cook, Curtin, Gordon, Left, McCombs, & Miller, 1987). 
Abstract issues may be hard for the public to comprehend and identify with, 
therefore making it difficult to attribute salience. This contrast between 
abstract and concrete issues is a significant one, yet it is often difficult to 
operationalize. After all, perceptions of issues in this manner are varied 
because they lie within the eye of the beholder. 

Another factor of notable importance is that of issue obtrusiveness. The 
obtrusiveness of an issue is the degree to which an issue forces itself into the 
direct experience of individuals (Yagade & Dozier, 1990). Past research 



16 



shows that the more real world experience an individual has with an issue the 
less likely agenda setting effects will be present (Winter & Eyal, 1981 ; 
Weaver, 1987). Thus, it appears that "the less information and sensitivity the 
public has to an issue prior to its treatment in the media, the more likely it is to 
be influenced by subsequent media reports" (Leff, Protess, & Brooks, 1986, 
p.31 1). Although this suggestion has been noted by several researchers, 
some have neglected to find stronger media effects for unobtrusive over more 
obtrusive issues (Demers, Craff, Choi, & Pessin, 1989). Erbring, Goldenberg, 
and Miller (1980) have even suggested that personal experience may 
enhance media effects. Certain issues may be primed by real world 
interaction, stimulating media interest. Demers (1996) found that personal 
interaction within a community actually predicted greater newspaper reading 
in that community. Whichever the case may be, there are indications that the 
obtrusiveness of an issue influences the public perception of that issue. 

The Real World Influence 

Another factor contributing to the formation of the public agenda involves 
the actual events that occur in our daily environment. There is the possibility 
that what researchers mistake for an agenda setting effect may essentially be 
nothing more than the media and the public responding in concert to real- 
world events (Iyengar, 1988; Mackuen, 1981; Funkhouser, 1973). This type 
of confounding variable, if not accounted for, can be very troubling for 
researchers. Behr and Iyengar (1 985) found that real world conditions 



17 



directly affected the economic issue of unemployment as well as 
complementing the media regarding other topics. Mackuen and Coombs 
(1981) proposed similar results with regard to economic issues. 

Funkhouser's (1973) study of several issues of the 1960's proposed a 
limited role of agenda setting by real world indicators. He found several 
cases in which the public agenda did not comply with real world indicators, 
yet it closely paralleled the media agenda. Beniger (1978) found that media 
coverage superceded social indicators in setting the public agenda. It has 
even been suggested that the media agenda bears no resemblance to the 
real world, yet has more powerful effects when determining the public agenda 
(Hester & Gonzenbach, 1994). Davis (1952) found that crime depictions in 
the Colorado area were not correlated with actual crime occurrences. It 
seems as though the media have the ability to create crime related fears with 
excessive coverage. Einsiedel, Salomone, and Schneider (1984) found that 
fear of victimization is better predicted by media exposure than by actual 
experience with crime. Individuals who pay more attention to televised news 
about crime are more fearful of crime related incidents (O'Keefe & Reid-Nash, 
1987). Fishman (1981) described the creation of an actual "crime wave" by 
the newspapers in New York City. This study indicated that the way in which 
the news was organized and selected affected the public's perception of the 
crime issue. Overall, it is evident that the real world measure of an issue 



18 



should be taken into account when researching who and what influence the 
public agenda. 

The media have a lofty responsibility to the American consumer. Some 
feel as though it is the job of the media to use their agenda setting influence 
to inform voters about candidates, issues, and images during elections 
(Weaver & McCombs, 1978). This information may promote awareness, but 
not necessarily knowledge. As Yankelovich (1985, p. 18) states, "It maximizes 
public awareness and public ignorance at the same time. To be excited 
about an issue but fail to think it through makes for the worst kind of citizen." 
The media's role may be one of civic mobilization (Shaw & McCombs, 1989). 
They help to focus attention on key issues in need of public action. It is 
unclear and uncertain what the ideal role of the media should include. 
Ultimately, it is up to the consumer to use the information they deem salient to 
take responsible action. 

The Influence of Policymakers 

The media are but one of the many factors that affect the agenda 
formation of the American public. Policy officials may also have an effect 
upon what the public decides to focus attention. This effect is often filtered 
through the media (Cohen, 1965). Roberts (1992) found that voting behavior 
may even be predicted by political issue attention in the media. Domke, Fan, 
Fibison, Shah, Smith, and Watts (1997) noted an agenda setting influence of 
candidate coverage on subsequent public opinion toward candidates in the 






19 



1996 presidential election. Clearly, politicians strenuously attempt to control 
their coverage by the media. They conveniently distance themselves from 
negative coverage and associate themselves with positive stories. This 
shows an attempt to reap the political benefits of agenda setting (Iyengar, 
1988). It is the desire of political hopefuls that positive media exposure will 
result in positive rapport with the American public. 

Many researchers have found that the president has a significant influence 
on the public agenda as well (Gonzenbach, 1992; Chaffee, 1978). Behr and 
Iyengar (1985, p.52) proclaim that "When the president speaks, the public 
listens." Political leaders are very powerful figures in the public agenda 
setting arena, but it is evident that not one factor alone shapes the agenda of 
the American public. The media, the president, the real world events, as well 
as personality and other unique factors all contribute to the formation of the 
public agenda. 

The Media Agenda 

In order to fully understand the agenda setting process, one cannot simply 
investigate the media's effect on public opinion. The formation of the media 
agenda itself must be accounted for. It is impossible to systematically study 
the rise of a given public issue without some understanding of these behind- 
the-scenes processes (Kosicki, 1993). An evaluation of the factors that 
promote the formation of the media agenda are key in understanding how 
particular issues are framed and offered to the public. 



20 



The Public's Influence 

It has been asserted that the media affect the public agenda, but the public 
may also have an effect on the formation of the media agenda. This idea of 
bi-directional agenda setting predicts that stories in the mass media may 
reflect a previous demonstration of interest in the public agenda. A number of 
researchers have found evidence in support of such a relationship (Watt & 
van den Berg, 1978, 1981; Behr & Iyengar, 1985; Weaver, 1987). 
Gonzenbach (1992) noted that the public agenda filters into the media 
agenda with regards to the drug issue. This suggests that consumers may 
very well affect media coverage. Some newspapers have even asked 
readers for input regarding coverage as a way to promote public journalism 
(Shepard, 1994). Rogers and Dearing (1987, p.571) concluded that there is a 
"two-way, mutually dependent relationship between the public agenda and 
the media agenda in the agenda-setting process." Erbring, Goldenberg, and 
Miller (1980) characterized this two-way relationship by emphasizing that the 
media agenda is set by the public in a gradual process whereas the media's 
effect on the public agenda is accelerated and more direct. The media seem 
to have a general idea regarding the interests of their audience. It would 
seem only natural then that coverage would reflect consumer demand in 
some cases. Distorted amounts of crime coverage have been shown to exist 
because of this need to pacify consumer demand (Graber, 1979). Stories 
might also reflect the media's desire to attract attention, therefore gaining 



21 



readers/viewers. Such stories play upon sensationalism, deviance, and 
tragedy. Ultimately, it is clear that the media are likely to cover issues they 
believe are of consequence to their audience (Carter, Stamm, & Heintz- 
Knowles, 1992). 

The Influence of Policymakers 

The public is certainly not the only factor influencing the media agenda. 
The president and various other policy elites affect the issues that appear on 
the media agenda as well. Walker (1977, p.433) stated that "for three safety 
related issues, the agenda of the U.S. Senate set the agenda for the New 
York Times." The most evident effects ultimately come from the president's 
agenda. Presidential activity can be said to cause a flurry of media activity 
(Behr & Iyengar, 1985). The president has the power to mobilize the media 
agenda merely by talking about an issue (Rogers, Dearing, & Chang, 1991). 
Wanta and Foote (1994) found this to be true for President George Bush with 
respect to issues on which he was an important source. 

Even those who have not yet made it to the White House may affect the 
mass media agenda. Iyengar and Kinder (1987) recognize that candidates 
for public office often go to extraordinary lengths to "make news." Roberts 
and McCombs (1994) found that political campaigns were instrumental in 
setting the media agenda during the 1990 Texas gubernatorial election. It 
appears that the media are at least somewhat influenced by the agendas of 
policy elites. 



22 



The Influence of Media Convergence 

Media agendas may also be formed in part by other media. Influential 
media have been identified as a major factor influencing the news media 
agenda (Danielian & Reese, 1989). This influence of media organizations on 
each other is termed "intermedia agenda setting" or "media convergence" 
(Danielian & Reese, 1989; Roberts & McCombs, 1994; Roberts, 1997). Too 
much sameness in media content may conflict with the American value of 
presenting diverse views and voices (Reese & Danielian, 1989). For this 
reason, media convergence is a topic of important study. 

McCombs and Shaw (1972), in their trailblazing study in agenda setting 
research, noted a striking similarity in related media coverage. Their results 
implied that a consensus on major news items might exist. McCombs and 
Shaw (1972), along with numerous researchers in the field, have asserted 
that the two leading sources for such an effect appear to be The New York 
Times and The Washington Post (McCombs & Shaw, 1972; Gans, 1979; 
Danielian & Reese, 1989; Shaw & McCombs, 1989). It has been said that 
when entertaining story ideas, editors will read the Times and Post to review 
the manner in which their editors cover a story (Gans, 1979). Ultimately, 
Gans (1979) notes that it is the "New York Times that is used as the final 
arbiter of quality and professionalism, across journalistic formats" (p. 1 81 ). 
Evidence of prior publication eliminates the need for an independent decision 
(Gans, 1979). As Sigal (1973) notes, "The consensible nature of news may 



23 



even impede the breaking of stories that lack corroboration from opinion- 
leading newspapers" (p.40). Consistency often implies accuracy. Breed 
(1980) describes the phenomenon of one newspaper leading others as a 
"dendritic" catalyst (p. 195). Just as the dendritic nerve cells carry impulses to 
the cell body, the influence of larger papers flows down to the smaller ones. 

More recently, Protess, Leff, Brooks, and Gordon (1985) found a similar 
effect. Their research revealed that investigative stories regarding rape in 
The Chicago Sun-Times promoted increases in subsequent stories on the 
same topic. Reese and Danielian (1989) detected media convergence in 
their investigation of media coverage of the drug issue in the mid-1 980's. 
Their research revealed that the New York Times took the lead in covering 
the cocaine issue, with the Washington Post following close behind. Other 
papers quickly fell in line and picked up the story and expanded upon it. 
Danielian and Reese (1989) note that such convergence in the media most 
often occurs "when a story is breaking, when coverage is at its peak, and 
when a story comes from a national or international source" (p.63). It is in 
these situations that news media cover a breaking story in a similar fashion. 

In Rogers, Dearing, and Chang's (1991) study of the AIDS issue, the mass 
media actually suppressed stories regarding the issue until it was picked up 
by the New York Times. Once the issue finally made it onto the media 
agenda, the national media analyzed were found to be highly correlated in 
their coverage (Rogers, Dearing, & Chang, 1991). Roberts and McCombs 



24 



(1994) discovered that even political advertisements exert an influence on 
newspaper and television coverage for political issues. They also confirmed 
an agenda setting influence of newspapers on subsequent television 
coverage. A replication of Roberts and McCombs study by Lopez-Escobar, 
Llamas, McCombs, and Lennon (1998) found similar results. Their results 
suggest an intermedia influence of newspaper coverage on the issues and 
attributes covered on television. Bichard (1999) noted a convergence of 
national media sources regarding the recent advancements and issues 
surrounding genetic cloning. Clearly, numerous media channels play a 
potential role in intermedia agenda setting. 

Interactive processes among various media, sources, and news 
organizations understandably influence the media agenda (Rogers, Dearing, 
& Bregman, 1993). Various peers in the newsroom or press pack might also 
have an effect. This is especially true when issues are given a cynical slant 
(Starobin, 1995). This type of peer pressure has been noted by ABC reporter 
John Quinones (Diamond, Accosta, & Thornton, 1987, p. 10) who made the 
statement that "sometimes we have a tendency to feed on one another." With 
the pressure of determining which stories are newsworthy, it may be easy to 
follow along with the ideas put forth by your peers. It takes a courageous 
journalist to challenge the norms created by the group (Shaw & McCombs, 
1989). 



25 



The Real World Influence 

Media coverage is at least partially determined by events in the real world 
(Funkhouser, 1973). Behr and Iyengar (1985) found this to be veritable of the 
television media agenda with regard to the issues of energy, unemployment, 
and inflation. Shaw and McCombs (1989) also detected that real world 
events affected story choices by journalists regarding the drug issue. Rogers, 
Dearing, and Chang (1991 ) discovered that many events precluded the mass 
media response to the AIDS issue. The most notable of those events 
included actor Rock Hudson's announcement that he had AIDS and the story 
of young Ryan White. Media coverage is said to have elevated at least ten- 
fold after these events took place. Rogers et al. (p.41 ), state that "The 
personification of the AIDS disease by White and Hudson has its primary 
impact in changing the meaning of the issue for media newspeople, who then 
responded with more attention to AIDS." This resemblance to the real world 
suggests that television coverage at least partially attempts to echo the actual 
state of the nation. 

The many factors contributing to the formation of the media agenda 
provide a useful picture of how issues are offered to the public. The ways in 
which these various factors contribute to how members of the public view 
issues will be an interesting topic of study for years to come. 



26 



The Policy Agenda 

Policymakers and governing officials also play a dynamic role in the 
agenda setting process. Items on the policy agenda or platform affect the 
political climate of the nation. How does an issue get on the policy agenda? 
Rogers and Dearing (1987) suggest two possible answers: 1 ) The public 
agenda, once set by, or reflected by, the media agenda, influences the policy 
agenda of elite decision makers, and, in some cases, policy implementations; 
and, 2) the media agenda seems to have direct, sometimes strong, influence 
upon the policy agenda of elite decision makers and, in some cases, policy 
implementation. It can be assumed that the Founding Fathers of our nation 
expected a substantial amount of constituency influence over the governing 
body. Burke's (Eulau & Wahlke, 1959) conception of this influence implies 
that the government should serve constituency's interests, but not its will. 
Many researchers have found such an influence. Page and Shapiro (1983) 
found a substantial causal correlation between public opinion and policy 
formation. Converse (1987) found that public opinion not only affected policy- 
making, but affected the presidential agenda as well. Gonzenbach (1992) 
found analogous results in his study concerning the drug issue. 

When considering policy agenda formation as a result of public concern, 
one might surmise that the sources communicating such concerns play a 
large role in the agenda setting process. In effect, the media may serve as 
the instrument through which public concern is crystallized in an attempt to 



27 



promote policy action. Shaw and McCombs (1989, p.1 18) clarify that "it is not 
the job of the news media to create social policy," but their agenda setting 
role provides an advantageous avenue through which the public and 
policymakers can be directed. Cohen (1965) may have captured the essence 
of this relationship best in his comparison of the media to the bloodstream in 
the human body. As our blood links the various parts of our body together, so 
does the media as it supplies us with political and intellectual nourishment. 

Numerous studies have indicated that the media have a significant effect 
on policy formation (Jones, 1976; Iyengar, 1988; Lambeth, 1978; Pritchard & 
Berkowitz, 1 993). Linsky (1 986) found that a high percentage of policy- 
makers admit such an influence from the mass media. In a study of media 
influence on prosecutors, Pritchard (1986) found that news media coverage 
was taken into careful consideration when deciding whether to engage in plea 
negotiations. Prosecutors were less inclined to plea bargain if a case had 
received news coverage. Policy allocation of funds can also be affected by 
the amount of media coverage on a topic. Pritchard and Berkowitz (1993) 
found that media coverage of the crime issue was a strong predictor of 
subsequent resource appropriation. The presidential policy agenda is also 
affected by the media agenda (Gonzenbach, 1992; Jablonski & Gonzenbach, 
1996). Wanta and Foote (1994) discovered that the President may use the 
media as a reference for which to decide which social problems are important 
to the American public. When operationalizing the policy agenda one must 



28 



be careful to distinguish between policy views and policy action. Protess, et 
al. (1987) found that the media affected attitude change of policymakers, yet 
policy action was more attributed to journalistic lobbying. Several other 
studies have also proposed that this active collaboration between journalists 
and policymakers often result in policy change (Cook, et al., 1983; Leff, et al., 
1986). The news media are believed by many people to be a reliable source 
of public concern. Policymakers often judge public sentiment directly from 
mass media suggestion. Overall, policymakers are very attentive to the 
public and media agenda when formulating the policy agenda. 

The Issue Attention Cycle 

Irrespective of how each of the agendas are formed, overall American 
sentiment rarely stays focused on any one issue for very long. This may be 
due to the vast amount of issues available to us. Miller (1965) suggests that 
the number of issues that can secure our attention at one time is relatively 
limited to seven, plus or minus two. Issues must then compete for our 
attention (Lippmann, 1922; Hilgartner & Bosk, 1988). This has been called the 
"zero-sum" principle (Zhu, 1992). The addition of any new issue on an 
agenda is at the expense of another issue. It has even been suggested that 
certain noteworthy events will create "killer" issues that have the power to 
move other issues off the agenda almost instantly (Brosius & Kepplinger, 
1 995). Lippmann (1 922) provides the example of a cue ball striking the end 
of a contiguous row of other billiard balls, it stops and another flies off the far 



29 



end of the row. In a like manner, the news, real world events, personality 
factors, and other factors, constantly have a dynamic competitive effect upon 
the public, media, and policy agendas. When an issue does secure a position 
on one of the various agendas, it is often short lived. Issues that remain 
prominent for a time inevitably wear thin and succumb to America's limited 
attention span (Zucker, 1978). This process appears to follow a cycle. 

Downs' (1972) termed it an "issue-attention cycle" and asserts that issues 
reflecting three specific characteristics are highly apt to follow such a cycle. 
First, a minority of the general public is suffering from the problem. Second, 
the distress incurred by such a problem results from actions that afford 
substantial benefit to a majority or a powerful minority of the population. And, 
third, the problem is not intrinsically stimulating in a sustained way. When a 
situation containing these three characteristics has captured the public eye, it 
is likely to move through the "issue-attention cycle." 

The cycle itself contains four stages, which may vary in duration, but most 
often occur in the sequence that follows. The cycle begins with the existence 
of a social concern or problematic issue. At this point the issue has not yet 
seized public attention so it is labeled the pre-problem stage. As a result of 
some dramatic series of events, or for other reasons, the public then 
experiences a sudden surge of attention towards the problem. This attention 
is coupled with alarm and euphoric enthusiasm about society's ability to 
"solve the problem" or "do something effective." This stage of the cycle is 



30 



entitled the discovery stage. Eventually there is a gradual realization that the 
problem cannot be solved easily and that solving it will be rather costly. This 
leads into the third stage of the cycle, which is labeled the plateau stage. 
People then inevitably become discouraged, threatened, and bored by the 
issue. Public desire to keep focused on the issue subsides. It is then that the 
issue reaches the final stage entitled the decline stage. The problem still 
exists, as well as various policy measures enacted in the previous stages, but 
public attention has migrated elsewhere from lack of interest. 

Issue Dynamics and Competition 

Following Downs' lead, several authors have since attempted to investigate 
the individual characteristics and issue dynamics that play a role in the 
competition for attention in different arenas. Competing trends are an 
important factor when analyzing issue attention. The unique environments 
within which issues compete are critical in understanding how trends are 
ultimately defined. It is useful to understand the competition that occurs 
between issues in the public environment. Ball-Rokeach and Defluer (1976) 
note the effects of issues in our environment with relation to a "dependency 
model." This model asserts that individual social settings will affect personal 
dependency on the media. For this reason, those in an environment lacking 
information pay more attention to issues covered by the mass media. This is 
most evident with respect to political issues. To the extent that an individual 
is dependent on media depictions of issues, they will pay more attention to 



31 



that information. Nimmo (1 978) went on to examine the types of political 
issues in the media that acquire increased attention. He found that new, 
unstructured, remote issues that appeal to emotional cues and depict people 
as issues were the most popular. He attributes such an effect to a shift in the 
public's political processing from being characterized as interested, 
knowledgeable, motivated, and democratic to a more disinterested, 
uninformed, and emotion-bound public. 

Various issues compete in the struggle for public attention. Hilgartner and 
Bosk (1988) propose that many social problems are defined differently with 
respect to their unique environments or arenas. Issues that survive are most 
often promoted by operatives with perceived credibility. The process can take 
years of promotion and interaction between arenas of discourse or it can 
occur almost instantly, but often the survival of one issue is at the expense of 
others. This dissolution factor increases competition. Some other variables 
that Hilgartner and Bosk (1988) propose to be determinants of attention 
include feedback and carrying capacity. Carrying capacity refers to the 
number of issues that can sustain salience at one point in time. They did not 
empirically test these assertions, but their guidance is undoubtedly 
appreciated and represented in future research. 

Neuman (1990) investigated the threshold of public and media attention to 
social issues. He examined the existence of a normally distributed logistic 
curve corresponding to a theory of public attention resulting from agenda 



32 



setting by the media. He described various stages necessary in the formation 
of public attention. These are reminiscent of some of the earlier mentions of 
collective behavior (Foote & Hart, 1953; Davison, 1958; Blumer, 1969). 
Neuman makes an interesting comparison to advertising in his attempt to 
explain lack of attention and issue dissolution. He proposes a "saturation 
effect." This occurs when a "unit increase in media attention no longer 
corresponds to an increase in public response" (Neuman, 1990, p. 162). 
Advertisers refer to such an effect as "ad wear-out" (Naples, 1979). When 
this "wear-out" or "saturation" occurs, the issue is often replaced by new, 
more interesting, and current issues. Neuman (1990) shows evidence to 
support the existence of a curvilinear logistic relationship between the volume 
of media coverage and the level of public concern. 

Brosius and Kepplinger (1995) also noted a correlation between media 
and public attention to issues. They investigated the effects of "killer" and 
"victim" issues. They note that an unexpected, surprising, or otherwise 
noteworthy event can kill existing issues by instantly moving into their position 
on an agenda. These "killer" issues are most likely to include information that 
is threatening, personally relevant, knowledgeable, or of symbolic value. 
Brosius and Kepplinger (1995) noted a "killer" effect from the issues of public 
security and energy. In their analysis, these issues were so dominant that 
they knocked several other issues off the public agenda. Their results 
confirmed the existence of both categories of issues, yet explained that the 



33 



process is very complex and other factors may also be contributors. With 
future study, expanded components of issue competition must be examined 
in order to arrive at predictions regarding the impact of social issues. 

Similarly, Zhu (1992) asserts that issue competition will always lead to the 
dissolution of some issue or issues. He proposes the existence of a zero-sum 
theory with relation to issue competition and attention. Zhu proposes that the 
promotion of any new issue causes the demise of another. Perhaps this 
occurs as a result of limited capacity. When the capacity is full, new issues 
must push old ones aside to gain an audience. McCombs (1995a) found 
similar results in their analysis of agenda-carrying capacity, agenda-diversity, 
and issue volatility. The study spanned a period of 40 years and showed no 
significant linear increase in the public agenda's carrying capacity. However, 
results did indicate that over the past 40 years there has been an increase in 
both agenda diversity and issue volatility. According to McCombs (1995a), 
the absence of an increased carrying capacity, coupled with increases in the 
diversity of issues and volatility, "leads to a faster rate of issue turnover on the 
public agenda." For this reason, issue operatives must strategically promote 
a limited amount of issues or attempt to "chunk" items together in a 
memorable way. Politicians have grasped this approach very effectively. 

The foundation of past research regarding patterns of attention and issue 
competition contributes to a more holistic view of the forces driving Downs' 
conceptualization of "issue-attention cycles." Issue attention and competition 



34 



are vital factors in the overall dynamic of agenda setting. American public, 
media, and policy agendas are all undoubtedly affected by this climate of 
attention and competition. The evolution of an issue may also have a direct 
effect on the media's chosen frame of reference. Further discussion of such 
an effect can be noted in the following section regarding a new dimension of 
agenda setting labeled framing. 

Framing 

Recently, the traditional agenda setting model has broadened to include a 
second level. The agenda setting model is no longer limited to the study of 
issue salience. As McCombs (1981; McCombs & Evatt, 1995) states, "in 
addition to providing cues about the salience of topics -objects, if you will- the 
mass media differentiate between the saliency of various attributes of these 
topics or objects" (p. 134). Attributes can be described as the qualities and 
characteristics that define an object (McCombs & Ghanem, 1998). McCombs 
(1992) further stresses that "both the selection of topics for the news agenda 
and the selection of frames for stories about those topics are powerful agenda 
setting roles" (p.8-9). The ability of the media to frame issues is now of great 
concern to researchers. By framing issues, the media not only tell us what to 
think about, but also how to think about it. The framing concept specifically 
focuses on the manner in which the construction of communication texts 
influences individual cognitions by selectively focusing on particular parts of 
reality while ignoring or downplaying other aspects (Shah, Domke & 



35 



Wackman, 1 996). Framing contributes to agenda setting theory by analyzing 
the content of issues in more depth. After all, as Kosicki (1993) notes, 
"agenda implies more than a simple list of topics" (p. 115). 

The theoretical underpinnings for framing research rests in both 
sociological and psychological literature. Although both use the term 
"framing" and share the idea that a frame is the perspective a person applies 
to define an issue or event, they differ in their application (Takeshita, 1997). 

The sociological perspective owes much to the work of Goffman (1974). 
His approach assumes that individuals classify and interpret life experiences 
in order to make sense of them. Social frames help us to "locate, perceive, 
identify, and label" information (Goffman, 1974, p.21). This perspective 
provides a vehicle in studying the mass media and issue framing. According 
to Gitlin (1980), media frames are "persistent patterns of cognition, 
interpretation, and presentation, of selection, emphasis and exclusion, by 
which symbol handlers routinely organize discourse" (p.7). He looked at the 
effects of framing with relation to media coverage of the 'students for a 
democratic society.' Gamson (1988) took up the study of media framing and 
the interaction among audience members and media content. He describes 
sets of issues in the media as "issue packages" (Gamson & Lasch, 1983; 
Gamson, 1988; Gamson & Modigliani, 1989). As Gamson and Modigliani 
(1989) describe, "A package has an internal structure. At its core is a central 
organizing idea, or frame, for making sense of relevant events, suggesting 



36 



what is at issue" (p.3). Gamson identifies five devices that signify the usage 
of frames: metaphors, exemplars, catchphrases, depictions, and visual 
images (Gamson & Lasch, 1983; Gamson & Modigliani, 1989). These frames 
influence the way audience members organize information. 

The most prominent views on framing in psychological literature began 
with the research of Tversky and Kahneman (1981 ,1982). According to their 
experimental studies, the cognitive perception of problems and evaluations 
produce predictable shifts of preference depending on how a message is 
framed. The representation of information affects judgment. This would be 
applicable to polling situations in which the wording of questions might exert 
influence on respondent choices (Takeshita, 1997). Iyengar (1991) has 
looked at this with respect to the news media and their depictions of issues 
such as poverty and crime. A series of experiments demonstrated that 
differences in the framing of such problems influenced people's perceptions 
of who was responsible. 

The concepts studied in framing research can be found in mass 
communication literature even before the term "framing" was applied. Many 
traditional media studies in agenda setting went beyond merely investigating 
issue salience (Benton & Frazier, 1976; Becker & McCombs, 1978; Weaver, 
Graber, McCombs & Eyal, 1981). Framing analysis has since been utilized 
by mass communication scholars, with many linking the sociological and 



37 






psychological origins to provide an integrated analysis of framing effects 
(Entman, 1993, Pan & Kosicki, 1993). 

Pan and Kosicki (1993) offer a model for analyzing news discourse with 
relation to sources, journalists, and audience members. The goal of this 
integration is to develop a way to "view news texts as a system of organized 
signifying elements that both indicate the advocacy of certain ideas and 
provide devices to encourage certain kinds of audience processing" (p.55-56). 
The structure of news discourse offers the following devices: syntactical 
structure, script structure, thematic structure, and rhetorical structure. A 
combination of what we learn about how the media operate in society and 
how media producers and consumers process information is necessary in 
order to fully understand the complete agenda setting process (Lasorsa, 
1997). The overlapping of the concepts from these two disciplines suggest 
that frames function as both "internal structures of the mind" and "devices 
embedded in political discourse" (Kinder & Sanders, 1990, p.74). 

Recent efforts in framing research have focused on defining and applying 
the term to mass communication and developing methods for analyzing 
framing attributes. Entman (1993) describes how framing offers an avenue 
for which to study the power of communication texts. In his assessment, 
frames serve the following functions in the communication process: "define 
problems, diagnose causes, make moral judgments, and suggest remedies" 
(p.52). Iyengar and Simon (1993) connect qualitative features of news 



38 



reporting on the Gulf Crisis to public opinion regarding the issue. They 
discuss the media's use of both "episodic" and "thematic" frames (p. 369). 
Episodic frames depict public issues through concrete and specific events. 
They usually contain visual appeal and include live on-the-scene coverage. 
Thematic frames place public issues in an abstract or more general context 
(Iyengar, 1991; Iyengar & Simon, 1993). They are more interpretive and 
focus on general outcomes or conditions. Results suggested that the 
television coverage of the Gulf Crisis was predominantly episodic and that 
"the pattern of episodic framing induced individuals to express greater support 
for a military as opposed to a diplomatic resolution of the crisis" (p. 366). 
Iyengar and Simon's frame distinctions, along with the inclusion of both media 
and audience perceptions, provides a useful example of how to integrate the 
psychological and sociological perspectives. 

Kosicki (1993) discusses framing research as a necessary addition to the 
traditional agenda setting paradigm because of its attempt to penetrate the 
"shell of the topic" (p. 112). Issues need the in-depth analysis that framing 
analysis can offer. He asserts that the future of agenda setting scholarship 
should go beyond issue saliency and into developing a more complete 
understanding of how various factors shape and engage public opinion. 

The many facets of framing analysis are discussed in length by Scheufele 
(1 996, 1 999). A systematic overview of framing research is given to offer 
insight into the task of clarifying and categorizing framing literature. A four- 



39 



cell typology of frames is suggested using frames as both dependent and 
independent variables for analyzing both the media and individuals. Possible 
operationalizations of each cell are discussed. Scheufele offers that "In order 
to fully explain and describe the influences and effects of framing in the field 
of mass media effects research, all four cells have to be part of an exhaustive 
concept explication on framing" (1996, p.9). A combination of methods that 
includes both the individual as well as the mass media will provide a more 
complete picture of the framing process. 

An illuminating description of this second level agenda setting can be 
found in Ghanem's (1997) recent depiction of framing research. She 
espouses McCombs and Evatt's (1995) idea that first level agenda setting is 
concerned with "objects" while second level agenda setting (framing) is 
concerned with the "attributes" of objects. Ghanem views framing analysis, 
not as a shift away from, but a way to build upon existing agenda setting 
research. In her words, "It is one highway linking up with another major 
thoroughfare" (Ghanem, 1997, p. 4). She discusses the original agenda 
setting hypothesis along with two additional framing hypotheses. The first 
hypothesis investigates the traditional first level effect of the media's attention 
to issues and subsequent public concern. The second hypothesis deals with 
the second level of agenda setting (framing) and "illustrates the influence of 
attribute salience on how people think about an object on the agenda" (p.4). 
The final hypothesis illustrates the influence of attributes in media coverage 



40 



on the salience of issues on the public agenda. This influence of second level 
media frames on first level public salience is termed "compelling arguments" 
(McCombs, 1996; Ghanem, 1996, 1997). Figure 1 provides a depiction of 
these relationships. 



First Level 



Second Level 



Media Agenda 

Objects 



Attributes 



1 . subtopics 

2. framing mechanisms 

3. cognitive elements 

4. affective elements 




Public Agenda 

~~ ► Objects 



► Attributes 



Figure 1: Two Levels of Agenda Setting (Ghanem, 1997) 



The examination of framing effects is best understood as a 
multidimensional representation. The interplay among various elements 
contributing to the dominant frame must not be overlooked (Hendrickson, 
1995). Ghanem (1997) refers to the metaphor of a "picture frame" in her 
conceptualization of framing attributes (p. 1 0). In order to measure these 
attributes, she offers four major dimensions: 1- the subtopics of a news item 
(what is included in the picture frame), 2- presentation or framing 
mechanisms (size and placement of the frame), 3- cognitive attributes (details 
of what is included in the frame), and 4- affective attributes (tone of the 



41 



picture). These four dimensions provide a comprehensive method for 
analyzing a variety of issues and topics (See Figure 1). 

Framing Dimensions 

The multidimensional aspect of issue frames takes into account the 
various layers that must be peeled away in order to analyze agenda setting 
effects. Each layer contains a unique insight into the overall framing process. 
The four dimensions are helpful in developing a basic understanding of that 
process. 

Subtopics are the first dimension for consideration when analyzing frames. 
The attributes in second level agenda setting (framing) are the independent 
variables. The categories used to name the attributes constitute the 
subtopics of an issue (Ghanem, 1997). Many of the traditional agenda setting 
researchers have grouped issues covered by the media to bring clarity to their 
analysis (McCombs & Shaw, 1972; McLeod, Becker, & Byrnes, 1974; Yagade 
& Dozier, 1990; Wanta & Foote, 1994; Brosius & Kepplinger, 1995). For the 
analysis of attributes in second level agenda setting, these sub-issues can be 
viewed as potential framing devices. The subtopics used to discuss an issue 
have great bearing on how that issue is perceived. Categorizing these topics 
provides a more concrete way for second level researchers to analyze 
agenda setting effects. 

The problem with relying simply on the categorization of sub-issues for 
framing analysis is the subjective nature of such choices. Choosing the topic 



42 



of a news item based on individual perceptions can lead to biased categories. 
The researcher comes up with a list of subtopics and then analyzes content 
and codes accordingly. Difficulty arises when there is little distinction 
between the content analysis in general and the examination of frames. For 
this reason, research must go beyond the examination and categorization of 
subtopics to include other dimensions. 

The second dimension includes an examination of the mechanisms used 
when covering issues in the media. Factors include the size, placement, 
photographs, pull quotes, and subheads used when covering a story 
(Ghanem, 1997). These all contribute to the prominence of the issue. 
Tankard (1991) terms these enhancing factors "framing mechanisms" (p. 15). 
These mechanisms provide a more in-depth picture of the topics covered in 
the media. 

Thirdly, in order to gain a complete picture of the framing of an issue, there 
must also be a consideration of the affective aspects of the story. The 
affective dimension deals with the public's emotional response to media 
coverage (Ghanem, 1 997). Use of story or chronicle narratives by journalists 
can have an effect on the public's emotional response and arousal (Donohew, 
1983; Ghanem, 1997). News values such as proximity, human interest, and 
personalization are also affective factors embedded in the framing process 
(Hall, 1981; Elliott, 1988; Fowler, 1991). The affective structure of coverage 
frames the issue by attaching a tone to the story. 



43 



The last dimension necessary for the investigation of the framing process 
is the cognitive element. This dimension "deals with the general cognitive 
categories that might shed light on whether the media and the audience are 
thinking about the problem in the same way" (Ghanem, 1997, p. 13). Studies 
have used such categories as conflict, thematic vs. episodic, causal 
responsibility vs. treatment responsibility, abstract vs. concrete, and cause vs. 
solution to describe the cognitive details of a topic (Klandermans & Sidney, 
1988; Edelstein, Ito & Kepplinger, 1989; Iyengar, 1989; Yagade & Dozier, 
1990; Iyengar, 1991; Edelstein, 1993; Iyengar & Simon, 1993). These 
cognitive categories move beyond topical categories by identifying the theme 
underlying the general coverage regardless of the topic. 

Used together, the four dimensions discussed provide a comprehensive 
picture of the attributes used in the media's framing of issues. Ghanem 
(1996) used all four dimensions to examine the framing of crime coverage in 
the media. The analysis focused on both first and second level agenda 
setting. Each of the dimensions included several variables based on previous 
literature. A total of 13 variables combined were used to represent the 
attributes of coverage. Significant relationships were found for each of the 
dimensions. Results indicated a correspondence between the four 
dimensions and public concern about crime. 

Framing research is relatively new to the agenda setting paradigm. It is a 
growing method of analysis and is sure to provide the field with rich insight. 



44 



In order for such efforts to succeed, a widely accepted conceptualization will 
have to emerge detailing the most productive way to categorize and analyze 
framing studies. As the investigation continues, so must the inclusion of the 
detailed influences of individual characteristics on framing effects (Miller & 
Wanta, 1996). The application of framing research is useful to a variety of 
issues and disciplines. Some of the most recent attempts to apply framing 
research have begun to emerge in the political arena. 

Framing in Political Discourse 

Framing analysis is especially useful when investigating political issues. 
Because of little direct experience with politics, people in modern society 
depend on the news media for information to understand their political world 
(Park & Kosicki, 1995). This provides the media with great power to focus 
and frame political reality. The political arena is replete with those attempting 
to frame issues in their favor. As Gans (1972) noted, "In any modern society 
in which a number of classes, ethnic and religious groups, age groups and 
political interests struggle among each other for control over the society's 
resources, there is also a struggle for the power to determine or influence the 
society's values, myths, symbols, and information" (p.373). With many 
political issues, there are proponents in constant struggle over how to define 
an issue. Schattschneider (1961) notes, "political conflict is not like an 
intercollegiate debate in which the opponents agree in advance on a definition 
of issues. As a matter of fact, the definition of the alternatives is the supreme 



45 



instrument of power. . . He who determines what politics is about runs the 
country, because the definition of alternatives is the choice of conflicts, and 
the choice of conflicts allocates power" (p.68). This power demonstrated in 
media choices provides a climate for provoking study. 

Entman (1991 ) examined the powerful effects of contrasting news frames 
employed by the media in the coverage of the KAL and Iran Air incidents. 
Political operatives played a role in shaping the contrasting frames. The 
words and images used in media coverage framed the KAL incident using a 
moral discourse, while confining the Iran incident into a technical one 
(Entman, 1 991 ). Public opinion polls confirmed that the American public was 
influenced by the discourse used in framing the incidents. 

Entman and Rojecki (1993) discuss the media coverage of the U.S. anti- 
nuclear movement from 1980-1983. They offer seven evaluative dimensions 
of the news message that are likely to affect how the movement will be 
framed. They include rationality-emotionality, expertise, public support, 
partisanship, unity, extremism, and power. Results indicated that coverage 
reflected the views of policy elites even when public support for the 
movement was abundant. Reese and Buckalew (1995) found similar results 
in their analysis of the framing of the Persian Gulf War. The media used 
"conflict, control, and consensus frames" (p. 47) to promote the heroic nature 
of the war and downplay any anti-war discourse. The valiant hero label 



46 



served policy interests by connecting the administration with the same heroic 
attributes. 

Park and Kosicki (1995) found the opposite to be true in the framing of the 
Iran-Contra affair. The frames used had a negative effect on the current 
administration. The media framed the issue as a "valence issue," defined as 
an issue with overwhelming consensus and little political debate (Stokes, 
1966, p. 172). They did not use "position" frames (p. 170), which would 
involve the intricate comparison of enacted policy to presidential expectations. 
The coverage of the Iran-Contra affair used valence frames by focusing on 
the administration's conspiracy in the secret dealing. Media frames prompted 
a simplistic judgment, thus having a negative influence on the public's 
presidential support. Research indicates a need to include policy agendas in 
the examination of framing effects in order to detect elite influence. 

Jasperson, Shah, Watts, Faber, and Fan (1998) investigated the media's 
framing of the federal budget deficit controversy. Agenda setting effects were 
examined at both first and second levels. Several frames were noted in the 
coverage between 1994-1996, and the "fight" frame had a significant impact 
on subsequent public opinion. Combined agenda setting and framing effects 
accounted for 92 percent of variance. Results further support the notion that 
agenda setting effects are enriched by the inclusion of frames utilized in the 
media's coverage of issues. Both contribute to the ability to predict the 
importance of political issues in the minds of voters. 



47 



In the political arena, individual voting behavior is vital. McCombs, 
Llamas, Lopez-Escobar, and Rey (1997,1998) investigated the framing 
effects of news articles and political advertisements on voter's perceptions in 
the 1995 elections held in Spain. They found a significant correspondence 
between the frames used and voters' substantive and affective perceptions of 
candidates. 

At the individual level, framing can be noted in the specific ways in which 
voters form their ideas and preferences. Frames can be thought of as a type 
of schemata in that they provide structure to everyday experience and 
facilitate the construction of meaning for individuals (Kosicki, 1993). How an 
individual interprets the media frame plays a role in subsequent decisions 
(Shah & Domke, 1995). lorio and Huxman (1996) found that individuals 
framed 1992 presidential election coverage by "linking, collapsing, and 
colorizing" their concerns. 

The media may shift the ground upon which elections are decided by 
priming citizens about the importance of particular issues and issue content 
(Iyengar & Kinder, 1987; Iyengar, 1990). Iyengar and Kinder (1987) argue 
that when ordinary individuals are faced with complex political issues or 
events, they are not able to conduct a thorough analysis based on prior 
knowledge because of limited information-processing capabilities. More 
often, people make judgments based on pieces of information that are easily 
retrieved from the top of their mind. 






48 



This idea originates from the psychological "cognitive priming theory" 
defined by Fiske and Taylor (1991 ) as the effects of prior context on the 
accessibility of information when processing new information. A 
conceptualization offered by Wyer and Srull (1986, 1989) describes 
information as stored in memory in the form of metaphorical storage bins. 
Primed categories are retrieved from a storage bin and then, rather than 
being replaced to its original location, the information is placed on top of that 
bin. This information is now more accessible in memory and will be used 
more easily in subsequent decision-making processes. Since the media 
provide much of what we know about political information, their choices in 
coverage can drastically affect the increased accessibility of certain topics. 
By "priming" these issues, individuals use them as criteria for evaluating 
political behavior (Price & Tewksbury, 1995). Therefore, the media not only 
affect the perceived salience of issues, but also the degree to which certain 
issues are used to judge political figures (Willnat, 1997). 

Iyengar, Peters, and Kinder (1982) investigated priming effects of media 
coverage on the evaluation of President Jimmy Carter. Their analysis 
showed a strong correlation between steady media coverage of defense 
issues and subsequent public evaluation of Carter's defense performance 
and competence. Iyengar (1 990) found similar results for those that 
evaluated President Ronald Reagan. He asserts that "the more prominent an 
issue in the national information stream, the greater the weight granted that 



49 



issue when individuals evaluate the president" (p. 171). Krosnick and Kinder 
(1990) found that judgements of Reagan's popularity became highly 
determinant on public support for the U.S. intervention in Central America 
immediately following the disclosure that funds from Iran arms sales had been 
used to finance the Contras. In short, evidence suggests that people use 
what they know to make judgements about political figures. If what they know 
is served to them by the media, then the increased accessibility or priming of 
such information will directly impact their subsequent evaluations. 

Iyengar and Simon (1993) found evidence that suggests that agenda 
setting, framing, and priming occurred in the news coverage of the Gulf Crisis. 
Television news coverage of the Persian Gulf conflict significantly affected 
Americans' political concerns and the criteria they used to evaluate President 
Bush. Krosnick and Brannon (1993) also looked at media coverage of the 
Gulf Crisis and subsequent effects on individual judgements. They found 
priming effects on presidential evaluation that were mediated by political 
knowledge, exposure, and interest. 

Shah, Domke, and Wackman (1996,1997) examined the effects of ethical 
textual frames versus material textual frames and found significant 
differences in the interpretations and subsequent decision making strategies 
of voters. These results suggest the news media's choices in selecting, 
emphasizing, and elaborating certain aspects of issues has a framing and 
priming effect on individual voter decision making. 



50 



If the media relay information regarding what issues are important, how we 
should feel about them, and what information we should access to judge 
others- then it is certainly worthwhile to study how and when these effects 
take place. What are the factors that contribute to such an affect? Many 
individual characteristics as well as issue dynamics are undoubtedly involved 
in this complex process. 

The media's choices for actively constructing news may vary according to 
different points during an issue's evolution (Berelson, 1948; Hall, Critcher, 
Jefferson, Clarke & Roberts, 1978; Lang & Lang, 1983; Linsky, 1986). Linsky 
(1986) concluded that media are most influential in the developing stages of 
an issue, while the problem and solution are still in "flux." The media's 
framing of issues may also depend upon the limited amount of information 
available and the resulting competition for agenda space (Zhu, 1992). 
Studies on issue competition and attention, such as Downs' issue attention 
cycle, would then provide useful boundaries for analyzing the media's framing 
of an issue. 

As Kosicki (1993) notes, "studying the attributes of the topics is another 
matter and has not often enough been explored by agenda setting scholars" 
(p. 104). This research seeks to use the agenda setting model to study media 
effects on public opinion in the context of public issues. Specifically, the 
campaign reform issue was chosen because of its controversial nature and 
recent intrigue with the American public. An analysis of public, media, and 



51 



congressional agendas regarding the issue will be conducted in order to 
reveal first and second level agenda setting effects. 

Research Questions 

The purpose of this research is to analyze the transfer of salience with 
respect to first and second level agenda setting. The issue used for the 
investigation will be political campaign reform. The traditional first level 
agenda setting hypothesis will be examined. A variety of research questions 
relating to the second level of agenda setting will also be investigated. See 
Figure 2 for a depiction of these relationships. The time frame for study will 
be determined by the cycle of attention described by Downs (1972) because 
the pattern of issue attention provides a useful perimeter within which to study 
the media's agenda setting patterns. The following research questions were 
developed in response to pertinent agenda setting literature. 



52 



<0 

e 
o 

0) CO 

O) c 
C 0) 
O O) 

o< 






coi2 




s 

ement 

lemen 


CO 


ub-lssues 
'aming 
echanism 
ffective El 
ognitive E 


(1) 


-•— ' 




CO u_ ^ < o 



t- CN 



(0 

■o 
c 

< 



3 
Q. 



to 

■o 
c 
(1) 

O) 

< 

(0 

CD 



CO 

"5 

CD 

o 





CO "* 



CO 

E » 

CO OT -S 

CD CD £ 

° m fi 

D rn CD 5 

CO •— ■— ."ti 

__ — ■*- J ^ 

T F o c 

-Q m CD O) 

C/> L_ < O 

r^ cn cd ^r 



CM CO "3- 



w 

<0 

_>, 

CO 

c 

< 

CD 
T3 

c 

co 
Q. 

£ 

to 
c 
g 

CD 

a: 

O) 

c 

ti 
CD 
CO 

CO 
T3 

C 
CD 

< 

1 

CD 

—I 

T3 

C 



CD 
CO 

c 
co 






> 
0) 



> 



u_ 






53 



First Level of Agenda Setting 



Media Agenda Public Agenda 

First Level Objects < ► Objects 



Figure 3: First Level Agenda Setting between the Media and the Public 

Research Question 1 

What is the direction of influence regarding the campaign reform issue on 
the media and public agendas (Figure 3)? This relationship will be 
represented by measuring cross-lagged correlation between the frequency of 
media coverage regarding campaign reform and the frequency of Internet 
discussion regarding the issue by the public. 

Rationale 

The assertion that the media affect what issues the public deem important 
is the most frequently studied relationship in the history of agenda setting 
research. Many scholars have found a strong relationship between these two 
agendas (McCombs & Shaw, 1972; Funkhouser, 1973; Beniger, 1978; 
Mackuen, 1981; Iyengar & Simon, 1993; Gonzenbach, 1996; Domkeetal., 
1997). Based on past research, I suspect that the media's coverage of 
campaign reform will be correlated with the public's perception of its 
importance. 



54 



Congressional 
Public Agenda Agenda 



First Level Objects < ► Objects 



Figure 4: First Level Agenda Setting between the Public and Congress 

Research Question 2 

What is the direction of influence regarding the salience of the campaign 
reform issue between the public agenda and the congressional agenda 
(Figure 4)? This research question will involve recording the frequency of 
public discussion on the Internet regarding campaign reform and 
congressional speeches referring to the issue of campaign reform. Cross- 
lagged correlations will be examined to establish the direction of influence. 

Rationale 

Much research has investigated the relationship between policymakers 
and the general public with little agreement. Some would argue that the 
public agenda, once set by the media, affects the policy agenda (Rogers & 
Dearing, 1987). Others have found a similar correlation between the public 
agenda and the presidential agenda (Converse, 1987; Gonzenbach, 1992). 
Many of the past analyses have used a variety of measures to represent the 
policy agenda including the president, Congress, and even court officials. It is 
difficult to arrive at a consensus with the use of different measures. For this 
reason, I have formulated a research question to investigate this relationship. 



55 



Congressional 
Media Agenda Agenda 



First Level Objects A ► Objects 



Figure 5: First Level Agenda Setting between the Media and Congress 

Research Question 3 

What is the direction of influence regarding the salience of the campaign 
reform issue between the media agenda and the congressional agenda 
(Figure 5)? This research question will involve recording the frequency of 
media coverage regarding campaign reform and congressional speeches 
referring to the issue of campaign reform. Cross-lagged correlations will be 
examined to investigate the direction of influence. 

Rationale 

Many scholars have asserted that the media have an affect on the policy 
agenda (Rogers & Dearing, 1987; Gonzenbach, 1992; Pritchard & Berkowitz, 
1993). The variety of measures used to represent the policy agenda has thus 
far led to inconsistent findings. For this reason, I have chosen to investigate 
this relationship with a research question. This analysis will use 
congressional speeches as a representation of the congressional agenda. I 
expect to gain insight into the direction of influence between the 
congressional agenda and the media agenda on campaign reform. 






56 
Second Level of Agenda Setting 





Media Agenda 


Public Agenda 


First Level 


Objects i^. 


Objects 


Second Level 


Attributes 4r^~ 

1. Sub-Issues 

2. Framing Mechanisms 

3. Affective Elements 

4. Cognitive Elements 


Attributes 



Figure 6: Second Level Agenda Setting between the Media's 
Attributes and the Public Agenda 



Research Question 4 

What is the direction of influence with regard to the media's coverage 
(attributes) of campaign reform and the salience attributed to the issue 
(objects) on the public agenda (Figure 6)? This will be investigated by 
analyzing the nature of the coverage of campaign reform along four 
dimensions (sub-issues, framing mechanisms, affective elements, & cognitive 
elements) and cross-correlating the results with the frequency of Internet 
discussion regarding the issue by the public. 

Rationale 

Ghanem (1996) asserts that the nature of coverage affects the amount of 
public concern with an issue. She found such to be true with respect to the 
crime issue. The attributes used in covering an issue in the media can have a 
powerful affect on the salience of an issue or topic on the public agenda and 
vice versa. This influence of media frames on the public agenda of objects 



57 



has been described as "compelling arguments" (McCombs, 1996). Although 
this influence has been noted, a sufficient amount of research to support a 
hypothesis has yet to emerge. For this reason, the analysis will be in the form 
of a research question. This study should provide yet one more step towards 
the establishment of a hypothesis. 



Media Agenda Public Agenda 

First Level Objects Objects 

Second Level Attributes < ► Attributes 

1. Sub-Issues 1. Sub-Issues 

2. Framing 2. Framing 
Mechanisms Mechanisms 

3. Affective Elements 3. Affective Elements 

4. Cognitive Elements 4. Cognitive Elements 



Figure 7: Second Level Agenda Setting between the Media and the Public 

Research Question 5 

What is the direction of influence regarding the nature of the media's 
coverage (attributes) of campaign reform and what the public thinks 
(attributes) about campaign reform (Figure 7)? This relationship will be 
evaluated by analyzing the cross-correlation between the sub-issues, framing 
mechanisms, affective, and cognitive attributes of media coverage and public 
discussion on the Internet. 



58 



Rationale 

This research question is the first example of a purely second level 
agenda setting or framing effect. If, as past literature suggests, the media are 
"stunningly successful" in telling the public what to think about (Cohen, 1963), 
can we assume that the media may also influence how we think about 
issues? Recent analyses have found that the way in which the media frame 
issues may have a significant affect on how the public accesses and 
considers such issues (Entman, 1991; McCombs et al., 1997, 1998; 
Jasperson et al., 1998). Realizing that the second level agenda setting effect 
has yet to be researched to the point of supporting a hypothesis, I have 
chosen to examine this relationship with a research question. I will examine 
the influence of both media and public framing along four dimensions (sub- 
issues, framing mechanisms, affective, and cognitive) utilizing methods used 
in past research. I expect to find a powerful relationship between the media's 
framing of these attributes with respect to the campaign reform issue and the 
frames utilized in the public's discussion of the issue. 



59 



First Level 
Second Level 


Public Agenda 

Objects 


Congressional 
Agenda 

Objects 
Attributes 

1. Sub-Issues 

2. Framing 
Mechanisms 

3. Affective Elements 

4. Cognitive Elements 


Attributes * ► 

1. Sub-Issues 

2. Framing 
Mechanisms 

3. Affective Elements 

4. Cognitive Elements 



Figure 8: Second Level Agenda Setting between the Public and Congress 

Research Question 6 

What is the direction of influence regarding the sub-issues, framing 
mechanisms, affective, and cognitive attributes of campaign reform on the 
public agenda and the congressional agenda (Figure 8)? This research 
question will involve the analysis of attributes in public discussion on the 
Internet and congressional speeches referring to the issue of campaign 
reform. Cross-lagged correlations will be examined to indicate the direction of 
influence. 

Rationale 

The connection between the frames utilized by policymakers and those 
used by the public is sure to be meaningful. Research has yet to determine 
the direction of influence with respect to second level agenda setting. I have 
chosen to analyze the framing of campaign reform in Congress and the public 
with a research question in hopes of gaining insight into this relationship. 



60 







Congressional 




Media Agenda 


Agenda 


First Level 


Objects 


Objects 


Second Level 


Attributes •* ► 


Attributes 




1. Sub-Issues 


1. Sub-Issues 




2. Framing 


2. Framing 




Mechanisms 


Mechanisms 




3. Affective Elements 


3. Affective Elements 




4. Cognitive Elements 


4. Cognitive Elements 



Figure 9: Second Level Agenda Setting between the Media and Congress 

Research Question 7 

What is the direction of influence regarding the sub-issues, framing 
mechanisms, affective attributes, and cognitive attributes for the campaign 
reform issue on the media and the congressional agendas (Figure 9)? This 
research question will involve the analysis of framing attributes utilized in 
media coverage and congressional speeches referring to the issue of 
campaign reform. Cross-lagged correlations will be investigated to infer the 
direction of influence for this relationship. 

Rationale 

While many scholars have recently investigated the relationship of media 
framing and public opinion, little research has focused specifically on the 
framing effects in the congressional agenda. This may be due to a lack of 
consensus regarding the representation of various policy agendas. This 
study will examine the relationship between the media's framing of campaign 



61 



reform and the frames used by the policymakers in Congress. I suspect that 
there will be a meaningful connection between the two. 

The following chapter will address each of the research questions 
mentioned and outline the specific methods that will be used to analyze them. 



CHAPTER 3 
METHODOLOGY 

Agenda setting studies have traditionally utilized a variety of methods. The 
most common include representations of the public agenda through polling or 
survey data, representations of the media agenda through content analysis of 
various media sources, and representations of the policy agenda through the 
analysis of presidential or congressional remarks (McCombs & Shaw, 1972; 
Funkhouser, 1973; Winter & Eyal, 1981; Shaw & McCombs, 1989; Cook & 
Skogan, 1990; Gonzenbach, 1996). Some have chosen to manipulate the 
variables experimentally (Iyengar, Peters & Kinder, 1983; Iyengar & Kinder, 
1987; Iyengar, 1989, 1991). The uses of a variety of methods have given the 
agenda setting paradigm added strength. This is especially true with respect 
to the first level or traditional agenda setting hypothesis. For the second level 
of agenda setting this remains to be seen. 

It is the purpose of this study to investigate both levels of agenda setting 
using some familiar methods as well as some utilizing the new media. This 
study will examine measures of the media agenda, the public agenda, and 
the congressional agenda for the issue of campaign reform. 

Content analysis will be used to gain unobtrusive quantitative measures 
for each of the agendas. This objective research technique offers scholars a 
replicable and accurate avenue through which to interpret communication 

62 



63 



texts. Frequency counts and coding techniques will be used to identify the 
first and second level agenda setting effects. 

The combination of these three agendas will add richness to the research 
design. As Kosicki (1993) urges, "each part of the process is incomplete and 
somewhat unsatisfying by itself, but by combining all three perspectives, the 
field can come closer to what a solid contemporary model of media influence 
ought to be" (p. 101). 

The following sections will discuss the chosen time frame for analysis and 
introduce the method for analyzing each agenda with respect to the campaign 
reform issue. 

The Issue Attention Cycle 

As mentioned in the review of pertinent literature, the time frame used for 
investigation of effects can have an enormous impact on the findings. For 
this reason, this study will attempt to analyze the issue of campaign reform in 
accordance with Downs' "issue attention cycle." In order to recognize the 
appearance of increased issue attention, the issue of campaign reform was 
recorded at six-month intervals in the New York Times and the Washington 
Post from 1 988 until 1 999. Figure 1 shows the result of the analysis. 

According to the coverage in these two prominent media outlets, the 
attention afforded the issue of campaign reform is on the rise. With a small 
peak seen in 1993 and a larger peak in attention beginning in 1996, the cycle 
of issue attention appears to be presently in a period of plateau. This study 



64 



c 
o 

CO 

c 

I 

I 





£ 
p 

I 



Q) 



o 



o 

CM 



O 
O 



O 
CO 



O 
CD 







A 




o> 






<^ 




a> 






^6> 




i 










CO 
00 










to" 
o 










Q. 

c 
o 






♦. 




CO 










c 








2 






A 

A 
A 


C 

o 
E 

X 

'35 


CO 
CO 

.c 
c 

CD 
CO 






A 


■o 
c 
o 
o 


1 
E 




L 


A 
A 


co 
w 
II 

£ 


c 
o 

>- 
i 






A 


C 

o 
E 


2 






A 

% 


X 

1 

II 
n 


E 

1— 

o 

CO 






A 


(0 


c 






.• 


O) 






A 
A 

\ 


>- 


CO 

Q. 

E 

CO 

o 








o 

B 
CO 






a. 




E 






\ 




CD 

> 
o 






\ 




o 

o 






X 




8> 

3 


1 1 


/ 


% 




CO 


1 1 






o o 

■* CM 


o 


% 







seuojs jo jequinN 



65 



will then take the most recent upswing in 1999 as an opportune time to 
investigate agenda setting effects. It is during the later portion of 1999 that 
the issue most recently took a marked upswing in media coverage. 

In order to capture the time order and direction of influence for the three 
agendas, the entire year of 1999 will be analyzed. The three agendas (public, 
media, & Congress) will be monitored weekly for the 52 weeks in 1999. This 
will allow the use of a time series analysis. The statistical technique will 
utilize an ARIMA model to analyze the longitudinal data for each of the 
agendas (Yaffee & McGee, 2000). This method removes naturally occurring 
trends over time before comparing one data set with another to provide a 
more accurate analysis (Box & Jenkins, 1976; McCleary & Hay, 1980). This 
technique has been used with frequency in a variety of longitudinal studies 
(Gonzenbach, 1996; Jablonski & Bichard, 1997; Corbett & Mori, 1999; 
Roberts, Wanta & Dzwo, 1999). 

Another advantage in using time series analysis is the ability to test for 
effects at several different time lags. The issue of time lag for effects is a very 
sensitive one (Chaffee, 1972; Winter & Eyal, 1981 ). If measurements are 
taken at a time lag that is too short, results will not capture the causal 
relationship. Measurements taken at a time lag that is too long may also be 
problematic. In this case, the effects may dissipate and therefore go 
undetected. 



66 



While many researchers suggest an optimal time lag of 4 to 6 weeks 
(Winter & Eyal, 1981), agenda setting effects have been known to vary 
depending on the measures used to represent agendas and the issues 
analyzed. Roberts, Wanta, and Dzwo (1999) found agenda setting effects for 
three issues in as little as 1 week when analyzing media coverage and 
Internet discussion. Others have found optimal lags for effects on the public 
agenda at up to four to six months (Stone & McCombs, 1981 ; Shoemaker et 
al., 1989). Wanta and Hu (1994) assessed agenda setting effects at several 
time lags and found that optimal lag times varied across different media. One 
to two weeks appeared to be optimal for television news, while newspapers 
had an optimal lag of three to four weeks. 

In the study most similar to the present analysis, where the public agenda 
is represented by Internet discussion and the media agenda represented by a 
content analysis of mainstream media sources, an optimum time lag of less 
than 1 week was noted (Roberts, Wanta & Dzwo, 1999). The reason for such 
quick effects appears to be the ease with which individuals have access to 
Internet forums. It is likely that individuals will immediately respond to media 
coverage when discussing issues on the Internet. 

Obviously, time lag can vary with each agenda being measured. Time 
series analysis allows for differences in time lag among separate variables. It 
is for this reason that the use of time series analysis in this study will provide 
the most accurate assessment of agenda setting effects. The following 



67 



paragraphs will explain the measures chosen to represent each agenda and 
how they will be examined in order to address the research questions in this 
study. 

The Media Agenda 

The analysis of the media agenda will be represented by stories in the 
New York Times and the Washington Post. They will be accessed using the 
Lexis/Nexis database (See sample in Appendix A). Several studies have 
indicated that these two media frequently coincide and even preclude the 
coverage of issues by other media in the country (McCombs & Shaw, 1972; 
Gans, 1979; Winter & Eyal, 1981; Danielian & Reese, 1989; Shaw& 
McCombs, 1989; Bichard, 1999). In order to validate this assertion, a weekly 
count of stories regarding campaign reform was recorded for these two media 
sources as well as three other major regional newspapers (The Boston 
Globe, The Houston Chronicle, & The Atlanta Journal and Constitution). The 
results are shown in Figure 1 1 . The figure indicates that these five 
newspapers follow a similar pattern of convergence on the issue. The use of 
the New York Times and the Washington Post should then provide an 
accurate depiction of the overall media agenda for the issue of campaign 
reform. Again, it can be noted that a surge of attention to the issue of 
campaign reform exists in the later portion of 1999. This further validates the 
decision to focus on the full year of 1999 to capture agenda setting effects. 



68 



o 
O 

c 
o 

i 





E 




o 




1 




•*-> 




A 




c 




a 




T3 




C 


1 


(C 


o 


— 


Q. 

c 
o 


E 

3 
O 

-J 


c 


■ 


1 


*-> 

c 
as 

s 


a; 


0) 


s 


1- 



w o 

II 
If 

if 

a> o 

Z I 

a> <D 

.c £ 
I- I- 




(0 



a) 
a: 

c 
g> 
'to 

a. 

E 

TO 

o 

«♦- 

o 

0) 
O) 
TO 

L- 

0) 

> 
o 

o 

i— 
0) 
Q. 
(0 
Q. 
(/) 

0) 

z 

a> 
o> 



3 



S8.uo)s jo JsquinN 



69 



Content analysis will be used to determine the media agenda for both first 
and second level agenda setting effects. For the first level of agenda setting, 
the content analysis will focus on the frequency of articles regarding the 
campaign reform issue. These will include any mentions of campaign finance 
reform or political campaign reform during the time period allotted. The 
Lexis/Nexis database will be used with the search string "campaign reform or 
campaign finance reform." Editorials will be eliminated as well as weekly 
summaries or articles written for regional publications. Once the articles are 
accumulated for the 52 weeks in the 1999 time period, they will be analyzed 
again for their relevance to the research project. Articles pertaining to 
campaign reform will be counted, whereas stories with a mere mention of 
campaign reform in the context of irrelevant discussion will be eliminated from 
the pool. The frequency count obtained will be used to represent the media 
agenda for the first level agenda setting analysis and to address research 
questions 1 and 3. 

To investigate the second level of agenda setting, the content of the 
articles chosen to represent the media agenda will be scrutinized and coded 
for sub-issues, framing mechanisms, affective elements, and cognitive 
elements. The unit of analysis for the second level agenda setting analysis 
will be each article. Coding will be based on those used by Ghanem (1996) in 
her analysis of crime attributes in the media. Sub-issue attributes will refer to 
the subtopics that emerge regarding campaign reform as well as political 



70 



candidate references or partisan associations. The framing mechanisms 
coded will include the length of the article, the location of the article, and the 
presence of illustrative elements. The affective dimension will refer to the 
positive, negative, or neutral tone of the article referring to campaign reform 
issues. This will represent the hostility or even optimism displayed in the 
article. The cognitive dimension will refer to the use of causes, solutions, or 
both to frame the issue of campaign reform. 

Upon completion, a random selection of 10% of the articles coded will be 
given to two additional coders for analysis to determine reliability (Holsti, 
1 969). These coders will be instructed on how to code each of the attribute 
dimensions. Intercoder reliability will be established to the degree that shared 
meanings emerge in the coding process. This examination of the second 
level of agenda setting effects will directly address research questions 4, 5, 
and 7. 

The Internet 

The Internet is not traditionally used as a research tool in mass 
communication analyses. It is still considered a new media and has just 
recently emerged in scholarly study. The vast exchange of information 
available through electronic bulletin boards, mailing lists, and moderated 
newsgroups provide many avenues to study on-line communication structure. 
This new dimension of public discourse has impacted society with more than 
100 million people now on-line (Weise, 1998). The impact of such an 






71 



enormous outlet for communication must be investigated in order to establish 
its placement in mass communication theory. 

The Internet is considered a mass communication tool in that it has the 
capability to address thousands of people, yet it can still function with respect 
to individuals one-on-one. The impact of such flexibility is sure to provoke 
communication research opportunities for years to come. The present study 
investigates on-line communication as a representation of the public agenda 
by studying USENET discussion messages. 

The USENET discussion groups are used by individuals to post a 
message about a given subject and then provide others on the Internet the 
opportunity to reply. This allows a public forum or debate to exist among 
individuals regardless of who they are or where they live. The USENET 
allows people to connect and transmit their views 24 hours a day at their own 
convenience. Users may discuss overtly or anonymously if they wish on a 
wide variety of issues. Some argue that the popularity of this on-line manner 
of discussion has resulted from the increased distrust and cynicism with 
traditional media outlets (Berger, 1995). Researchers have found that the 
interactivity of USENET groups provides a key avenue through which to 
monitor the opinions of the public (Bobbitt, 1995; Dzwo, 1998). Research has 
even noted that political discussion in USENET groups interact with dialogue 
similar to that of traditional social groups (Hill & Hughes, 1997). 



72 



The integrity of the data accessed from on-line discussion groups 
depends heavily on the sincerity of the comments posted, therefore a profile 
of Internet users is necessary in order to establish a better understanding of 
the sample chosen for this investigation. The applicability of the findings to 
the public in general also depends on the similarities and differences detected 
between Internet users and non-users. 

Of the more than 100 million people on-line, over 60 million reside in the 
United States and the predominant language of the Internet is English 
(Sampat, 1998). The use of the Internet appears to be dominated by 
"boomers and young adults" (Russell, 1998). A current survey of Internet 
users by Mediamark Research (1999) asserts that 23% of adults under age 
55 have been on-line in the past 30 days, while only 5% of adults 55 and over 
recorded usage of the Internet within the past 30 days. The same survey 
found that 50.8% of Internet users are men, 61.7% of users are married, 48% 
have children living at home, 75.1% have college education, 70.8% have jobs 
in the general management or service industry, and 33.1% have an income of 
less than $50,000 a year. 

A similar survey of American Internet users was conducted by the Luntz 
Research Company in 1997. This analysis was useful in that it offered a 
comparison of those who are Internet users with those that are non-users. 
The poll depicted users or the "connected" as those that exchange email at 
least once a week and use at least one additional technology regularly. 



73 



Additional technologies included cell phones, laptop computers, 
beeper/pagers, and home computers. The results suggest that 20% of those 
"connected" watch between 1 1 -20 hours of television in a given week, with 
1 9% of those "unconnected" watching 1 1 -20 hours of television a week. As 
for obtaining news information, 24% of the "connected" look to newspapers 
and 20% look to cable television. For the "unconnected" the figures are 34% 
and 25% respectively. With respect to political involvement, 36% of the 
"connected" said that they always vote and 47% of the "unconnected" said 
that they always vote. Confidence in the political system was also similar for 
both the "connected" and "unconnected" with 36.2% and 42% respectively 
saying that they have "a lot of confidence in democracy". Many of the figures 
represented in the survey for both groups are strikingly similar. As the 
number of Internet users in America grows, the profile of users to non-users is 
becoming more and more alike (Edmondson, 1997). Although Internet users 
are still considered a distinct group, the bias is diminishing with each year. 

Although Internet usage is prevalent, it cannot be asserted that users 
participating in USENET discussion groups are identical to the average 
American. The political involvement levels are certain to be heightened for 
these individuals if they are making the additional effort to assert their views 
in on-line discussion. This study, then, looks for a more powerful effect than 
traditional agenda setting research. The study will examine effects taken to 
the behavioral level. In other words, the use of such a measure actually 



74 



works against finding agenda setting effects. The investigation of the 
communication patterns on this new medium are well worth the added 
pressures in analysis. 

The Public Agenda 

As indicated, the investigation of the public agenda will be represented by 
public discussion regarding campaign reform on the Internet. The traditional 
representation of the public agenda with polling data was avoided because of 
a lack of consistency in polling questions (See Table 1 ). This lack of 
consistency was noted in campaign reform question samples pulled for the 
1999 time frame. The wording of poll questions must have similarity in order 
to accurately measure changes in salience over time. Also, the campaign 
reform issue is not widely mentioned in 1 999 as one of the most important 
problems in the nation when compared with problems such as crime, 
education, and the economy. The Internet, on the other hand, provides an 
avenue through which users can discuss a wide variety of issues. For this 
reason, the use of Internet discussion groups will be utilized to provide a 
depiction of public opinion regarding campaign reform in 1999. The use of 
on-line USENET discussion groups to examine political behavior is 
advantageous because it is accessible, free of charge, and can be instantly 
downloaded (Groper, 1996). 



75 





Table 1 : Polling data samples for 1999 




Date 


Question 


Percentage 


Sample 


Polling 
Organization 


1/1999 


Which of the following 
issues should receive the 
highest priority from 
Congress? (listing of 
various issues) 


3% answered 

campaign 

reform 


1,000 
National 
registere 
d voters 


American 
Viewpoint National 
Monitor Survey 


1/1999 


What do you think are the 
two most important issues 
that the federal government 
should address? 


1% answered 

campaign 

reform 


1,011 

National 

adults 


Hart and Teeter 

Research 

Companies 


2/1999 


What do you think are the 
two most important issues 
that the federal government 
should address? 


2% answered 

campaign 

reform 


924 

National 
registere 
d voters 


Opinion Dynamics 
Poll 


5/1999 


Which one of the following 
issues should receive the 
highest priority from the 
Congress this year? (listing 
of various issues) 


2% answered 

campaign 

reform 


1,000 

National 

adults 


American 
Viewpoint National 
Monitor Survey 


7/1999 


What one issue would you 
like to hear the presidential 
candidates talk more about 
next year? 


1% answered 

campaign 

reform 


1,200 

National 

adults 


Pew Research 
Center 


10/1999 


Suppose Congress passes 
campaign finance reform 
legislation. Do you think 
there would be a positive 
effect, no effect, or don't 
know? 


Positive- 43% 
None- 44% 
Don't know- 
13% 


1,003 

National 

adults 


Wall Street 
Journal Poll 


12/1999 


Which candidate do you 
most associate with the 
campaign reform issue? 


George W. 

Bush- 16% 

Steve Forbes- 

7% 

John McCain- 

18% 

Al Gore- 15% 

Bill Bradley- 

8% 

Other/Don't 

know- 36% 


1,073 

National 

adults 


Pew Research 
Center 


12/1999 


What is the single most 
important issue to you in 
deciding who to support for 
president? (listing of 
various issues) 


1 % answered 

campaign 

reform 


1,006 

National 

adults 


Washington Post 
Poll 



(Refer to Appendix B for complete list) 



76 



A search engine called Deja.com was chosen to access Internet 
discussion (See sample in Appendix C). Deja.com (http: //www.deja.com) 
archives over 40,000 on-line USENET discussion forums from as far back as 
March, 1995. Past analyses have supported the usage of on-line text to 
represent the public agenda (Dzwo, 1998; Roberts, Wanta & Dzwo, 1999). 

The unit of analysis represents each message retrieved using the search 
terms "campaign reform or campaign finance reform" for the 52 weeks in 
1999. Although some argue that a message thread must include at least two 
response messages in addition to the original posted message (Rafaeli, 
1 988), this study analyzes every message under a thread as an independent 
unit as long as it relates to the issue of campaign reform. For this analysis, 
the discussion postings are being recorded as measures of public opinion and 
therefore do not necessitate responsive interactivity. 

Once the pool of messages are accumulated, they will be scrutinized for 
their relevance to the research project. Messages containing discussion of 
campaign reform will be counted, whereas mere mentions of the issue in the 
context of irrelevant discussion will be discarded. The frequency of postings 
will be used in the investigation of first level agenda setting effects in an effort 
to address research questions 1 and 2. 

For the second level agenda setting analysis, the USENET discussion 
messages will be examined and coded for sub-issues, framing mechanisms, 
affective elements, and cognitive elements. The unit of analysis will be each 









71 



individual message. Sub-issues will refer to the subtopics that are identified 
in the messages regarding campaign reform as well as political candidate 
references or partisan associations. Framing mechanisms will include the 
length of the message and the use of quoted references. The illustrative 
elements and the prominence in location will not be used for this agenda. 
The affective dimension will refer to the overall positive, negative, or neutral 
tone of the posted message. This does not refer to evaluations regarding 
campaign reform, but rather the overall tone of their language. The cognitive 
dimension will refer to the use of causes, solutions, or both to frame the 
content of the message. 

Upon completion, a random selection of 10% of the coded messages will 
be given to two additional coders for analysis to establish reliability (Holsti, 
1 969). These coders will be instructed on how to code each of the attribute 
dimensions in on-line text. This examination of second level agenda setting 
effects will specifically address research questions 4,5, and 6. 

The Congressional Agenda 

The congressional agenda will be represented by the content analysis of 
Congressional Records. These records contain textual representations of 
speeches given by congressional members on the House and Senate floor. 
They will be accessed using the Lexis/Nexis database (See sample in 
Appendix D). Cook and Skogan (1990) found such measures to adequately 
represent the congressional agenda. 



78 



The Lexis/Nexis search term "campaign reform or campaign finance 
reform and 1 minute" will be used to obtain speeches in the 1 -minute format 
for the 52 weeks in 1999. The 1 -minute speech format was chosen for 
analysis to eliminate the extremely lengthy debates that likely refer to 
campaign reform briefly in the context of other discussion. Accumulated text 
will then be analyzed to further determine their relevancy to the research 
project. Speeches simply referring to campaign reform, but focusing on other 
topics, will be deleted from the pool. First level agenda setting will be 
examined by the frequency of records containing appropriate discussion of 
political campaign reform or campaign finance reform. This data will be used 
to investigate research questions 1 and 2. 

For the analysis of second level agenda setting, the chosen speeches will 
be scrutinized and coded for sub-issues, framing mechanisms, affective 
elements, and cognitive elements. The unit of analysis will be each individual 
speech. Sub-issue attributes will refer to the subtopics that are identified in 
the speeches regarding campaign reform as well as references to political 
candidates or partisan platforms. The framing mechanisms utilized will 
include the length of the speech and the person giving the speech. Once 
again, the use of illustrative elements and location are inappropriate for this 
agenda. The affective dimension will refer to the overall positive, negative, or 
neutral tone of the speech. The cognitive dimension will refer to the use of 
causes, solutions, or both in framing the text of the speech. 



79 



A random selection of 10% of the chosen speeches will be given to two 
additional coders for examination to establish reliability (Holsti, 1969). This 
examination will ultimately provide answers applicable to research questions 
3 and 4. 

Coding Categories 

A description of the coding categories used to investigate the attributes for 
the second level agenda setting analysis is necessary in order to provide an 
accurate picture of how the comparisons will be made. Each of the three 
agendas represented will be discussed with reference to specific coding 
categories (See Code Sheets in Appendix E). The following nine topics were 
chosen based on interviews with prominent researchers in the field of political 
science. These topics were chosen to represent the sub-issue attributes for 
the issue of campaign reform: 

■ Freedom of speech: This topic refers to any discussion of money as 
speech or 1 st amendment rights. 

■ Soft Money: This topic refers to discussion regarding unregulated 
contributions from special interest groups, unions, or corporations. 

■ Candidate advertising: This topic refers to discussion of political 
advertising, free commercial time, or express advocacy of a candidate. 

■ Incumbency Advantage: This topic refers to the increased resources 
incumbents have access to- examples include franking privileges, 
gerrymandering, and the like. 

■ Contribution Limits: This topic refers to discussion regarding limiting 
campaign contributions to candidates running for office and includes limits 
on personal contributions made by candidates. 



80 



■ Expenditure Limits: This topic refers to the discussion of limiting the 
amounts that candidates can spend to run a political campaign. 

■ Political corruption: This topic refers to mentions of a pervasive 
American cynicism with the corruption of the democratic process. 

■ Specific Legislation: This refers to specific discussion regarding 
campaign reform legislation or bills before Congress (example- FECA, 
1971 or Shays/Meehan). 

■ Reform proponents: This refers to the discussion of individuals that 
support campaign reform such as McCain, Feingold, Shays, or Meehan. 

The framing mechanisms chosen were duplicated from those used by 

Ghanem (1996) to denote prominence in the display of information. The 

following descriptions will be used to offer a descriptive analysis: 

■ Length: This will include the number of words in the text used for all 
three agendas. 

■ Location (only for media agenda): Articles will be coded for placement 
either on the front main page, front page of inside section, or elsewhere 
inside the newspaper. 

■ Illustrative Elements (only for media agenda): Articles will be coded for 
the presence of photographs or other graphics. 

■ Quotes (only for public agenda): Messages from the Internet discussion 
groups will be coded for the number of quoted references. This will be 
used in the absence of location or illustrative elements to provide an 
indication of prominence. 

■ Speaker (only for congressional agenda): Speeches on the House and 
Senate floor will be coded with respect to the partisanship of the speaker. 
This will be used in the absence of location or illustrative elements to 
provide an indication of prominence. 

The affective attributes for this analysis will denote the tone of the text. 

Past research has used similar depictions for the affective dimension 

(McCombs, Llamas, Lopez-Escobar, & Rey, 1997; Lopez-Escobar, Llamas, 



81 



McCombs, & Lennon, 1998; Huckins, 1999; McCombs, Lopez-Escobar, & 
Llamas, 2000). The following descriptions will be used to categorize the 
positive, negative, or neutral tone of the material analyzed: 

■ Negative: This refers to discussion or text that is generally engaging in 
a negative or judgmental tone, hostile or conflictual, using attacks or 
aggressive language, pessimistic or sarcastic evaluations of the topic and 
those involved (word examples include: corrupt, disgust, frustration, 
chaotic). 

■ Positive: This refers to discussion or text that is generally making 
judgments in a positive or optimistic tone, stressing fairness and 
acceptance, presenting views using logical and non-inflammatory 
language, with a calm and sensitive demeanor (word examples include: 
agree, change, compromise, helpful, benefit). 

■ Neutral: This refers to discussion or text where there is an absence of 
judgment; statements are of a more factual nature; or non-evaluative 
reports of related events, gatherings, and meetings. 

The cognitive dimension for this analysis will refer to the use of causes, 

solutions, or both to frame discussion of campaign reform. These cognitive 

attributes were replicated from past research (Iyengar, 1989, 1991; Entman, 

1991; Rucinski, 1992; Ghanem, 1996). The codes for this dimension are as 

follows: 

■ Cause: This refers to attributions of causal responsibility, identification 
of who/what is responsible for problematic campaign financing or the 
absence of meaningful reform. 

■ Solution: This refers to attributions of treatment responsibility, mentions 
of who/what offer remedies, and how the problem should be alleviated. 

■ Both: This refers to attributions/identification of both the causes and 
solutions necessary to combat the problem of campaign financing. 



82 



The following chapter will provide detailed results of the content analysis. 
The sample will be discussed as well as the findings for each of the agendas. 
Each research question will be addressed with respect to both first and 
second level agenda setting effects. 






CHAPTER 4 
RESULTS 

This chapter provides a description of the sample used for analysis. A 

discussion regarding the frequency distribution for each sample will be offered 

as well as statistical analyses of each time series. The analysis will then 

address the results for each of the research questions offered in the previous 

chapter. 

Descriptive Results of Content Analysis 
The Media Agenda 

Articles were retrieved from the New York Times and the Washington Post 
using the Lexis/Nexis database. The search string "campaign reform or 
campaign finance reform" was utilized to facilitate the search. A total of 260 
articles were retrieved for the year 1 999. These articles were then analyzed 
to assess their relevance to the research project. Weekly summaries or 
regional publications were eliminated from the sample pool. Articles merely 
mentioning "campaign" or "reform" without association were discarded. A 
valid sample of 132 articles remained. Ironically, an equal number of articles 
were utilized from each source with 66 coming from the New York Times and 
66 from the Washington Post. This pool of articles is smaller than anticipated, 



83 



84 



but an acceptable number considering the small amount of attention afforded 
the issue by the media. 

The content analysis of articles revealed media coverage of a variety of 
sub-issues for the topic of campaign reform. Table 2 provides percentages 
for each of the primary sub-issues covered during the time frame of the study. 
The most frequently covered topic involved discussion of reform legislation 
(33.3%). The second most frequently covered campaign reform topic focused 



Table 2: Campaign Reform 


Sub-Issues Covered 


by the Media 


Subtopic 


Percentage 


Reform Legislation 


33.3% 


Reform Proponents 


22% 


Soft Money 


11.4% 


Free Speech 


8.3% 


Contribution Limits 


5.3% 


Political Corruption 


4.5% 


Candidate Advertising 


3% 


Incumbent Advantage 


2.3% 


Expenditure Limits 


1 .5% 


None of the Above 


8.3% 


Total 




132 



on specific reform proponents (22%). These two sub-issues alone accounted 
for over 55 percent of the discussion by the media regarding campaign 
reform. 

The frequencies of affective and cognitive elements in the media's 
coverage of campaign reform are depicted in Table 3. The affective tone of 
the articles was most often negative (42.4%). The articles maintaining a 
positive or neutral affective tone were fairly even split at 27.3 percent and 
23.5 percent respectively. The cognitive elements coded referred to causal 



85 



solution, or frames utilizing both causal and solution combined. Causal 
attributes indicated responsibility for the problem. Solution attributes referred 
to remedies for the problem. Combined usage refers to those utilizing both 
causal and solution attributes together to frame the issue. The most frequent 
cognitive element used by the media was the combined causal/solution frame 
(35.6%). 

The media's use of framing mechanisms varied. Table 4 depicts the 
number of words most commonly used to discuss an article regarding 
campaign reform. The average number fell at 61 7 words per article. 



Table 3: 


Affective and Cognitive Attributes in Media 




Coverage of Campaign Reform 


Affective 




Cognitive 


Attributes 


Percentage 


Attributes Percentage 


Negative 


42.4% 


Causal & 35.6% 
Solution 


Positive 


27.3% 


Solution 27.3% 


Neutral 


23.5% 


Causal 21 .2% 


None 


6.8% 


None 15.9% 


Total 


132 



Table 4: Number of Words used 

in Media Coverage of 

Campaign Reform 



# of Words 


Percentage 


Under 100 


5.3% 


100-300 


17.4% 


301-500 


16.7% 


501- 700 


22% 


701-900 


22% 


901-1,000 


4.5% 


Over 1 ,000 


12.1% 


Mean- 617 


Std. Deviation- 416 


Total 


132 






86 

The location and use of graphics were also noted as framing mechanisms 
in the media's depiction of the campaign reform issue. Table 5 offers 
percentages for each element coded during the time frame analyzed. The 
location of articles regarding campaign reform were most often inside the 
paper with a very small percentage of articles located on the front page or 
front inside section (5.3%). 

Graphics were used sparingly when discussing campaign reform. Over 86 
percent of the total sample did not offer illustrations, photos, or other 
graphical material (See Table 5). 

Table 5: Location of Articles & 
Use of Graphics 




Location Percentage 




Inside Elsewhere 94.7% 
Front Page/ Front 5.3% 
Inside Section 


Graphics Percentage 


Illustrations / Graphic 
Inserts 9.8% 
Photographs 3.8% 
None 86.4% 


Total 132 


The Congressional Agenda 

Speeches made in the House and Senate were retrieved from the 
Congressional Record using the Lexis/Nexis database. The search string 
"campaign reform or campaign finance reform and 1 minute" was used to 
obtain the sample. One minute speeches were used to represent the 
member's views on campaign reform. The search yielded a total of 85 



87 



speeches for the 1 999 time frame. These were then analyzed for their 
relevance to the research project. Mere mentions of campaign reform in the 
midst of irrelevant discussion were discarded. A valid sample of 62 speeches 
remained. All of the 62 speeches chosen for the sample came from the floor 
of the House of Representatives. This is not entirely surprising given that the 
1 -minute debate format is more prevalent in the House than in the Senate. 
The issue of campaign reform also had yet to reach the Senate for debate in 
full force. The sample pool is again smaller than expected, but is acceptable 
given the low frequency of congressional discussion for campaign reform in 
relation to other topics for the 1999 time frame. 

The content analysis of the speeches in the sample revealed a relatively 
limited use of sub-issues for the campaign reform topic. Table 6 shows a 
breakdown of the primary sub-issues mentioned and their percentages. 



Table 6: Campaign Reform 

Sub-Issues In 

Congressional Speeches 



Subtopic Percentage 



Reform Legislation 66.1% 

Political Corruption 17.7% 

Soft Money 6.5% 

None of the Above 9.7% 



Total 62 



Reform legislation was by far the most dominant frame used in the speeches 
(66.1%). The political corruption frame is a distant second with a percentage 
of 17.7. 



88 



Affective and cognitive elements were detected in most of the speeches 

analyzed. Table 7 offers the percentages for each of the framing attributes. 

The affective tone of the speeches was evenly split with 46.8 percent 

displaying a negative tone and 46.8 having a positive tone. The cognitive 

elements utilized were predominantly a combination of causal and solution 

frames (54.8%). 

Table 7: Affective and Cognitive Attributes in 

Congressional Speeches Regarding 

Campaign Reform 



Affective 




Cognitive 




Attributes 


Percentage 


Attributes 


Percentage 


Negative 


46.8% 


Causal & 
Solution 


54.8% 


Positive 


46.8% 


Solution 


25.8% 


Neutral 


0% 


Causal 


17.7% 


None 


6.4% 


None 


1 .7% 


Total 






62 



Framing mechanisms for the congressional speeches simply referred to 
the number of words used and the partisanship of the speaker. Table 8 
depicts the percentages for each element. The majority of speeches 
contained between 100 and 200 words (67.7%). The average length for a 
speech was 1 92 words. 

The partisanship for speakers on the campaign reform issue was 
overwhelmingly Democrat. Republican and Independent speakers only 
accounted for 17.7 percent of the sample analyzed (See Table 8). 



89 



Table 8: Length of Speeches & 
Speaker Partisanship 



# of Words 


Percentage 


Under 100 


3.3% 


100-200 


67.7% 


201-300 


24.2% 


Over 300 


4.8% 


Mean 192, St.dev. 96 




Partisanship 


Percentage 


Democrat 


77.5% 


Republican 


16.1% 


Independent 


1.6% 


None of the Above 


4.8% 


Total 


62 



The Public Agenda 

The analysis of the public agenda for the campaign reform issue was 
represented by public discussion on the Internet. USENET discussion groups 
were retrieved using the search engine Deja.com. The initial examination 
retrieved over 12,000 messages for the campaign reform issue in 1999. The 
search string was then modified to include only the words "campaign reform" 
within 30 characters. This eliminated discussion of the two words "campaign" 
and "reform" without a relevant association. The search then revealed a total 
of 11 8 messages for the 1 999 time frame. These were initially analyzed for 
their relevance to the research objective and a valid sample of 81 messages 
was chosen for further examination. Unfortunately, before the messages 
could be downloaded for further analysis, the search engine went out of 
business. The archives were sold to another on-line source and are no 
longer available to the public. The number of articles per week will still be 



90 



applied for the traditional agenda setting results, but the framing analysis will 
have to be postponed until the discussion archives are accessible. 

Research Questions 

A total of 275 representations of the campaign reform issue were coded 
for the combined analysis of the media, public, and congressional agendas. 
An intercoder reliability check produced a reliability coefficient of .89 for these 
coding decisions. 

ARIMA Modeling 

Time series analysis was used to examine the 7 research questions in this 
study. Each series containing the 52 weekly data points for 1 999 was first 
modeled using ARIMA statistical techniques. Autoregressive integrated 
moving averages (ARIMA) time series analysis essentially estimates three 
parameters and can be expressed as ARIMA (p.d.q). The modeled 
processes are autoregression (AR), integration (I), and moving averages 
(MA). An ARIMA model predicts a value in a response time series analysis 
as a linear combination of its own past values and error (Box & Jenkins, 1976; 
SPSS, 1988). 

The first parameter included in the (p,d,q) ARIMA model is autoregression 
(AR) and can be estimated by the strength or dependency of the correlation 
between directly adjacent values in a series. The second parameter for 
estimation includes differencing in order to strip off the integration (I) of the 









91 



series. This is necessary for any series exhibiting trend-like behavior, 
nonstationarity, or unsteady changes from one observation to the next. The 
third and final parameter for modeling is the moving average (MA) component 
and is characterized by the disturbance and persistence of random shock in a 
series. This is the degree to which each value in a series is determined by 
the average of the current disturbance and one or more previous 
disturbances. 

The three parameters in an ARIMA (p,d,q) process are closely related and 
must be chosen carefully in order to build the most appropriate model for a 
given data set. This requires careful identification, estimation, and diagnosis 
until a satisfactory model is achieved (SPSS, 1988). 

Examining autocorrelation (ACF) and partial autocorrelation (PACF) 
functions for each series is the first step in identifying your parameter choices 
(p,d,q). Once estimates are chosen, you can look to the reported t ratios for 
the modeled series as well as the Akaike Information Criterion (AIC) and 
Schwartz Bayesian Criterion (SBC) goodness of fit statistics. Significant 
values suggest a successful model. Lastly, you must plot the autocorrelation 
(ACF) and partial autocorrelation (PACF) of the resulting error series for your 
chosen model. This series contains the residuals from the model and should 
produce randomly distributed nonsignificant (according to the Box-Ljung 
statistic) correlations resembling white noise. This permits a final diagnosis 
regarding the appropriateness of your model. If the suitability of the chosen 



92 



model appears questionable at any time during the identification, estimation, 
and diagnosis process you must repeat the steps until a satisfactory fit is 
achieved (Yaffee & McGee, 2000). 

For the current study, each univariate series of data was modeled until an 
acceptable fit was achieved according to the above criteria. The results 
include a table for each series to demonstrate the chosen ARIMA model 
parameters and coefficients. Once the fitted series were built for each data 
set, a cross-correlation analysis was calculated in order to test the specific 
relationships addressed in the research questions for this examination of the 
campaign reform issue. The results are presented in tables detailing the 
cross-correlation coefficients at a variety of time lags and their corresponding 
significance. 



First Level of Agenda Setting 



Media Agenda Public Agenda Congressional 

Agenda 

First Level Objects -#- -► Objects 

Objects 



Figure 12: First Level Agenda Setting Relationships 

Research questions 1-3 addressed the direction of influence regarding the 
campaign reform issue in media coverage, public Internet discussion, and 
congressional speeches (Figure 12). This examination focuses on the 



93 



traditional first level agenda setting relationship by investigating the salience 
of the issue. Figure 1 3 depicts the frequency distribution for each agenda in 
1999. It can be noted that the three agendas appear to be related in their 
focus on the issue, but the direction of influence for these relationships is not 
apparent without further statistical analysis. 

The nature of the three time series was first explicated using the ARIMA 
modeling technique. Table 9 displays the ARIMA model parameters chosen 
for the media, public, and congressional series. These series address only 
the first level agenda setting analysis. The model for the media agenda 
indicated a differenced, first-order moving average process: (0,1,1). The 
model for both the public and congressional agendas indicated a differenced, 
first-order autoregressive process: (1,1,0). All three coefficients appeared to 
significantly represent well-fitted models for these series. 



Table 9: Public, Media, and Congressional Agendas 
ARIMA Analysis of Univariate Series 



Series 


ARIMA term 


Coefficients 


t ratio 


P 


Public 


(1,1,0) 


(f>1 = -.66 


-2.70 


.009 


Media 


(0,1,1) 


c|)1 = .93 


11.23 


.000 


Congress 


(1,1,0) 


(|>1 = -.50 


-4.05 


.000 



Once the models for each univariate series were determined, bivariate 
relationships were examined using cross-correlation analysis. Table 10 
provides a depiction of the public, media, and congressional correlations from 



94 




1.9 
6fr 
Lf 
SV 
& 
t* 
6£ 

ze 

9C 

ee 



T— 

E 

k_ 

o 
>•— 

a: 

c 
g> 

CO 

Q- 

E 
re 
O 

2 

O 
0. 

c 
o 



IE 


<0 




re 




TJ 


6Z 


c 

(U 




O) 


« 1 


< 


i 


re 

c 


93 


o 








1 


ez 


£ 




o> 




c 


12 


o 




O 


61 


o0 




o 


a 


-Q 




3 




Q- 


St 


. 




re 




T3 


St 


0) 




^ 


tt 


re 




1- 


6 


co 




T— 


L 


re 




3 




D) 


9 


LL 


e 




i 





ssajBuoo ui seipeads jo 'saBessaui i3N3Sn 'sapivie jo jaqwiiN 



95 



a -7 to +7 week lag. As Table 10 indicates, there are several significant 
correlations. The public agenda precedes the congressional agenda with 
significant correlations at 2 to 3 weeks. The congressional agenda precedes 
the media agenda with significant correlations at 4 to 6 weeks. It is not a 
surprise then that the public agenda precedes the media agenda with its 
strongest correlation at a total lag of 5 to 7 weeks. The public agenda is also 
moderately correlated with the media agenda at a lag of 0,1 ,3 and 4 weeks. 

Research question 1 specifically addressed the relationship between the 
salience of the campaign reform issue on the media agenda and the public 
agenda. The results indicate a strong agenda setting effect of the public 
on the media with the strongest correlations of .41 and .42 when public 
discussion precedes media coverage by 6 to 7 weeks respectively (p<05). 
Moderate agenda setting effects were also discovered for the public as a 
leading indicator of the media with significant correlations when the public 
discussion precedes media coverage at 1 and 3 to 5 weeks (p<05). The last 
significant correlation of .30 between the two agendas occurs at weeks 
when the agendas are considered synchronous (p< 05). 

Research question 2 specifically addressed the direction of influence 
regarding the salience of campaign reform between the public agenda and 
the congressional agenda. A significant agenda setting effect of the public on 
Congress was discovered with a correlation of .32 and .33 when public 









96 



O 
2 



CO 
CN 






in 



CN 



+ 



CD 

+ 



V) 

O 
2 



o 



CO 

o 



+ 



CD 

+ 



CO 
CD 
T3 

c 

< 

CO 

c 
g 

CO 
CO 

CD 

l_ 

O) 

c 
o 
O 

XJ 

c 

CD 

CD 

TD 
CD 



O 
Ql 



CO 

c 
o 

JO 

I 

i_ 

O 

O 

T3 
CD 
O) 
O) 

JD 
i 

CO 
CO 

2 

O 



.Q 
CD 

H 



CO 
CN 



in 

CN 



CM 



in 

CN 



CD 
CN 



CO 

o 



1^ 
o 



CO 

o 



+ 



+ 



CO 

+ 



CN 

+ 



CN 

o 



CN 

o 



in 

+ 



+ 



CO 

+ 



CN 

+ 



in 
o 



o 

3 

D 
Q. 



10 

a 

o 

c 

e 



! 



o 

CO 



CO 
CN 



CN 



* 
CN 



* 

CN 

CO 



CO 
CO 



CO 
CO 

O) 

c 
o 
o 



CN 
CN 



CN 
CO 



CO 
CO 



CO 



CO 
CN 



CN 
CN 



CO 

O 

c 
8 

■8 



CN 



CO 



in 



CD 



CD 



CO 
CN 



CN 
CO 



CO 



CN 



CO 



o 

CO 



in 



CD 



CN 


q 

i 




CO 




V) 


CO 


O) 


U 


c 


CD 


o 


^ 


O 



1 



in 

CN 



D) 
CO 



CO 





CD 
CO 



in 
o 

v' 

CL 

* 



i 

c 
c 

TJ 
CO 
iP. 

CO 
(0 

<n 

c 
o 
O 
i_ 
o 
o 

Q. 
CD 



CO 

.c 

■•^ 
CD 
O 

c 

CD 

2 
CD 
N 

O) 

C 

TJ 

8 
2 

Q. 
(0 
O) 

ra 

CO 
(0 

c 
g 
*■*-» 

I 

o 
o 



s 



c 
(0 

CD 
O 



97 



discussion on the Internet precedes congressional speeches by 2 to 3 weeks 
respectively (p< 05). 

Research question 3 specifically examined the direction of influence 
regarding the salience of the campaign reform issue between the media and 
congressional agendas. A significant agenda setting effect of the Congress 
on the media was noted with a correlation of .32 and .34 when congressional 
speeches precede media coverage by 4 to 5 weeks respectively (p<05). 



Second Level of Agenda Setting 





Media Agenda 


Public Agenda 


First Level 


Objects i, 


r Objects 


Second Level 


Attributes 4r^* 

1. Sub-Issues 

2. Affective Elements 

3. Cognitive Elements 


Attributes 



Figure 14: Second Level Agenda Setting between the Media 
Attributes and the Public Agenda 



Research question 4 examines the nature of the media's coverage of 
campaign reform and the salience of the issue in public Internet discussion 
(Figure 14). This analysis addresses both first and second level agenda 
setting by positing a relationship between framing attributes and overall object 
salience. Figure 15 gives a complete depiction of the 1999 frequency 
distribution for the public agenda and a variety of media attributes concerning 
campaign reform. 



98 



o 

is 

Q. 



0) 

> 



S 



ro 



a 

1 

■ 
■ 
■ 

8 



CO 

ra 







E 
o 

<D 

a: 

c 
2> 
re 

a. 

E 
re 
O 

re 
o 

jp 

o 

a. 

c 
o 

re 

T3 

c 
a> 

< 

3 
Q. 

a> 



c 
re 

CO 

a> 

-9 

1 

O) 

c 

1 

re 

u_ 

Li- 
re 
t5 
a> 



re 

-t— < 

o 

H 



o 


in 


o 


M 


o 


m 


o 


If) 


lO 


* 


■* 


CO 


CO 


CM 


CM 





saBesssiu 13N3SD J° JaqiunN 9 sauiaij jo JdqiuriN 



99 






Sub-issues represent the most numerous media framing attributes with an 
average of 8.7 frames utilized per week. Affective and cognitive attributes in 
the media are represented with a moderate frequency of 2.3 and 3.0 frames 
per week respectively. The media's uses of framing mechanisms are not 
presented because they are not numerical in nature. The discussion of 
framing mechanisms has been thoroughly addressed in the previous section 
on descriptive results. The public agenda for the issue is represented by 
Internet discussion with an average of 1 .6 messages per week. Statistical 
analysis of each series will reveal possible agenda setting effects. 

Each time series was first defined using ARIMA modeling techniques. 
Table 1 1 presents the parameters chosen to fit the ARIMA models for the 
public agenda and the three series of media attributes. The model used for 
the public agenda indicated a differenced, first-order autoregressive process: 
(1 ,1 ,0). The model for both the sub-issues and cognitive media attributes 
indicated a differenced, first-order moving average process: (0,1 ,1 ). Finally, 
the model for the affective media attribute series indicated only a first-order 
moving average process: (0,0,1 ). All four coefficients are significant below 
the .05 level and appear to represent appropriately fitted models for these 
series. 



100 



Table 1 1 : Public Agenda and Media Framing Attributes 
ARIMA Analysis of Univariate Series 



Series 


ARIMA term 


Coefficients 


t ratio 


P 


Internet Agenda 


(1,1,0) 


<|)1 = -.66 


-2.70 


.009 


Media Subtopic 
Attributes 


(0,1,1) 


$1 = .73 


7.03 


.000 


Media Affective 
Attributes 


(0,0,1) 


<|>1 = -.28 


-2.07 


.043 


Media Cognitive 
Attributes 


(0,1,1) 


<|.1 = .79 


7.84 


.000 



Once the univariate models were determined for each series, bivariate 
relationships were examined using cross-correlation analysis. Table 12 
reports the overall relationships among the agendas. Cross-lagged 
correlations are offered at -7 to +7 week lags. The analysis indicates several 
significant correlations. 

Research question 4 specifically addressed the possible agenda setting 
relationship between the media's use of attributes and the salience of the 
campaign reform issue on the public agenda. The results indicate several 
strong associations. 

The public agenda appears to affect the media's use of subtopic attributes 
with the strongest correlations of .49, .50 and .46 when increased frequency 
of public discussion regarding campaign reform precedes the number of 
framing sub-issues utilized by 5 to 7 weeks respectively (p<05). Smaller but 



101 









CD 
CN 


q 


CO 


+ 










CO 




1 








CO 

CD 






o 












•*— » 






o 


* 


-<* 


CO 


CD 




_Q 






u» 


T— 


o 


q 


+ 




"co 

CO 


CO 






00 

q 


q 


q 


in 

+ 




I 


1 








1 








E 


.o 






h- 


in 


CD 


+ 











q 


q 


q 




D) 

C 


o 
















T3 


c 
















CO 


'e 






CO 


CN 


i-> 


CO 







2 






q 


q 


q 


+ 




CO 


LL 
















CO 


(D 
















■o 


T3 






q 


CO 

q 


q 


CN 

+ 




c 

CD 
CO 
CO 


T3 
















O 


C 
















iS 


CO 






00 


CD 


r-~ 


T — 




3 


CD 






q 


CN 


q 


+ 




3 


T3 


















c 




52 












CD 


< 

O 




5 












.C 


.9 

Q_ 


o 

c 

8 

•8 




* 
CN 




O 




*-< 
■4— • 

CO 

x: 

4— • 


n 




c 












0) 


Q_ 




£ 


* 

00 
CN 


00 


CN 


t— 




o 

c 

CD 


O 
















T3 


CO 
















o 


c 
















i_ 


g 






eo 


■>a- 


* 
CO 

co 


CNI 





N 


CO 
















CO 


CD 
















C 


i_ 


















i_ 
















■D 


o 
o 






* 

q 


o 

CN 


« 

ID 
CO 


CO 




8 



TJ 
















i_ 


CD 
















Q. 


O) 
















CO 


CO 

1 






* 

C7> 

q 




* 
00 

eo 


^}" 




O) 

12 -o 


CO 
















*- 


CO 
















10 c5 

CO o 


o 
















O 






* 


* 

CN 


* 


m 




II 








■<t 


CO 


TT 


i 




CN 
















iS co 


t - 
















^ 



CO 




M 


* 

o 

35 


o 

CN 


* 
CO 


CD 




b g> 

o £ 


h- 




i 


i 

CD 

w 

o 


O 


* 

CD 
> 


i 




|c5 

li 

CO c 

§ ■■ I 

H -C 






■ 


CO o 

CD "3 


> 

CD Jt 


si 


CO 

_J 




K z H 


1 




|2co 


S< 


SO . 







102 



still significant correlations also exist with the public Internet discussion 
preceding by 1 to 4 weeks (See Table 12). 

The public agenda also appears to affect the media's use of affective 
attributes with a correlation of .32 when the increased frequency of public 
Internet discussion of campaign reform precedes the number of affective 
frames utilized by 5 weeks (p< 05). A significant correlation of .29 is also 
apparent at weeks when the public agenda and the media's use of affective 
attributes are considered synchronous (p<05). 

The public agenda again emerges as a leading indicator of the media's 
use of cognitive attributes with strong correlations of .47, .49, and .47 when 
public discussion precedes the media's number of cognitive frames by 5, 6, 
and 7 weeks respectively (p<05). Smaller yet significant corellations can 
also be noted when the public agenda precedes by 2 to 4 weeks (See Table 
12). Overall, the public agenda appears to exert a somewhat powerful 
agenda setting effect on the media's usage of framing attributes most notably 
regarding the number of sub-issues and cognitive frames utilized. 



103 



Public Agenda Media Agenda Congressional 

Agenda 

Second Level Attributes < ► Attributes 



Attributes 

1 . Sub-Issues 

2. Affective Elements 

3. Cognitive Elements 



Figure 16: Second Level Agenda Setting Relationships 

Research questions 5-7 addressed the direction of influence regarding the 
nature (attributes) of the media's coverage, public discussion on the Internet, 
and congressional speeches for the campaign reform issue. This is the only 
purely second level agenda setting analysis in this study (Figure 16). As 
noted previously, an analysis of the framing attributes reflecting the public 
agenda will be omitted due to the fact that the USENET discussion archives 
are no longer accessible. For this reason, research questions 5 and 6 will 
remain unanswered until further analysis is permitted. The sub-issue, 
affective, and cognitive framing attributes used to address the issue will still 
be analyzed for the media and congressional agendas. This will directly 
examine research question 7. 

Before investigating the possible relationships between the two remaining 
agendas, each series of framing attributes were remodeled using the ARIMA 
statistical technique. The dimensions of the framing attributes will be 
analyzed separately with respect to each agenda so as to simplify the display 



104 



of results. Once the results for each dimension are explained, an overall 
analysis of the media and congressional framing attributes will be given to 
address research question 7. 

Sub-issue dimension. The first dimension for examination will be the 
primary sub-issues revealed in the media's coverage and congressional 
speeches regarding campaign reform. Primary and secondary attributes 
were coded distinctly so as to offer a greater depth of analysis for sub-issue 
framing effects. 

Table 13 depicts the ARIMA model parameters chosen for each series of 
primary sub-issue attributes in the media's coverage of campaign reform. The 
three primary sub-issues were soft money, political corruption, and reform 
legislation. All three models represent a significant fit. 



Table 13: Media Primary Sub-Issue Attributes 
ARIMA Analysis of Univariate Series 



Series 


ARIMA 
term 


Coefficients 


t ratio 


P 


Soft Money 


(1.1,0) 


4>1 = -44 


-3.47 


.001 


Political 
Corruption 


(0,0,1) 


+1 = -.65 


-5.80 


.000 


Reform 
Legislation 


(1,0,1) 


<|>1 = -.68 & -.93 


-4.08 & -9.02 


.000 & .000 



The ARIMA parameters for each series of the congressional primary sub- 
issue attributes are provided in Table 14. The three primary issues analyzed 
are also soft money, political corruption, and reform legislation. These were 



105 



chosen because they are the only three issues that were covered at the 
primary level by both the media and congressional agendas. 



Table 14: Congress Primary Sub-Issue Attributes 
ARIMA Analysis of Univariate Series 



Series 


ARIMA term 


Coefficients 


t ratio 


P 


Soft Money 


(1.1,0) 


<|>1 = -.49 


-3.97 


.000 


Political 
Corruption 


(1,1,0) 


<|>1 = -.48 


-3.87 


.000 


Reform 
Legislation 


(1,1,0) 


<|>1 = -.49 


-3.92 


.000 



After the models for the univariate series in each of the two agendas were 
determined, bivariate associations were examined using cross-correlation 
analysis. Table 15 gives an overall depiction of the relationships between the 
primary sub-issue framing attributes utilized by the media and Congress. 
These are provided at -7 to +7 week lags. Multiple significant correlations 
are evident. 

The results in table 15 indicate a moderate agenda setting effect of the 
congressional attributes on the media's attributes for the soft money frame. 
The positive correlation of .33, .42, and .31 occur when congressional framing 
attributes for the issue precedes media attribute coverage by 3, 4, and 5 
weeks respectively (p<05). This relationship suggests an agenda setting 
effect of the congressional framing agenda on the media framing agenda. 



106 



- 




(0 

o 


CM 


1^ 
q 

1 


q 


+ 




CD 






"o 


"* 


CD 


"j- 


CD 

+ 




O 






u: 


«^ 


q 


q 




-4— » 








1 


r 


1 






O 


















CD 


















CL 








■<r 


CD 


N- 


in 

+ 




CO 








1 


q 


q 
i 




CD 


















.c 


















*^ 


CO 
CD 






l" 


CN 

q 
r 


CN 

q 


+ 




CD 


















T3 


jQ 
'C 






in 


CD 

q 


ID 

q 
i" 


CO 

+ 




CD 

' E 

CD 

SZ 


CO 
















*— » 


c 
E 






lO 

o 

1 


q 


ID 

q 


CN 

+ 




CO 

CD 


CD 
















CD 


i_ 
















^~ 


LL 
O 






ID 
CN 


q 


* 


+ 




CO 
CO 

CD 


Q. 




(/) 




■ 








i_ 


O 


CO 


3 












O) 


s 

t" 


CO 
O) 

c 


o 

c 

8 


« 

CO 

CO 


in 
q 


CO 

q 


o 




C 

o 
o 

CD 


£-5 

CC CD 


o 

o 


I 


CD 


ID 


O) 






SZ 
*-* 


E ^ 




CO 


q 


q 


q 


*J" 




CO 


£e 








■ 


i" 






sz 


ons of 
ress a 






CD 

q 


ID 

q 
i* 


CN 

q 
i 


CN 




o 

c 

CD 
T3 


CD C 






• 

CO 


o 


ID 


CO 




O 


13 

o 






q 


CN 


*-[ 


1 




CD 
















N 


O 
CD 






* 

CN 


* 
O 

q 


■<r 


"<f 




c 
8 


c» 
















CD 


CO 

JD 

i 






« 
q 


* 

ID 


a> 
q 


in 




Q. 
CO 


CO 
















C» 


CO 

o 
















_CD 


l_ 






o 


t— 


a 






*H$ 


O 




CD 


q 


i" 


q 


CD 




CD 
CO 


in 




CD 












C 


T— 




O 


h- 


CN 


O) 






o 


CD 

-Q 




1 


i* 


q 
i 


q 


i 




CO 
CD 


CD 


















H 






CD 

c 


c 
o 


c 
o 






o 
o 








o 

o 


If 

o o 


E* 

is 

V c? 


to 




C CD 

gco 

~ T3 








CO 


Q-O 


a: _j 


_l 




CJ>b 


















CO CD 










CD 








£ ^ 










1 








X o'-b 










2 








« a zl5 



107 



Another significant correlation of .36 exists for the two agendas on soft money 
at weeks when the agendas are considered synchronous (p< 05). 

The results indicate a very strong correlation of .80 for the political 
corruption frame when congressional framing attributes precede media 
attribute coverage by 4 weeks (p< 05). A moderately strong correlation of .45 
is also noted for the political corruption frame when the congressional 
attributes for the issue precede media attributes by 5 weeks (p<05). This 
relationship again implies that the direction of agenda setting influence runs 
from the congressional framing agenda to the media framing agenda. 

Results indicate a significant relationship between the two agendas for the 
reform legislation sub-issue frame as well. A correlation of .41 is detected 
(Table 15) when the media's attribute coverage for the issue precedes 
congressional attributes by 1 week (p< 05). This is the only significant 
relationship in which the direction of influence depicts the media framing 
agenda as a leading indicator of the congressional framing agenda rather 
than vice versa. This may suggest a two-way relationship. 

Secondary sub-issues were also coded with respect to campaign reform 
on the media and congressional agendas. These issues were not necessarily 
the dominant focus, but were mentioned with enough frequency to merit 
analysis. Eight sub-issues were noted that received secondary emphasis on 
both agendas. The following issues are those chosen for comparison: free 



108 



speech, soft money, candidate advertising, contribution limits, expenditure 
limits, political corruption, reform legislation, and reform proponents. 

Table 16 provides the ARIMA parameter choices and the significance 
figures for each series of secondary sub-issue attributes used by the media. 



Table 16: Media Secondary Sub-Issue Attributes 
ARIMA Analysis of Univariate Series 



Series 



ARIMA 
term 



Coefficients 



t ratio 



Free 
Speech 


(0,0,1) 


fl=-.56 


-4.74 


.000 


Soft Money 


(1,1.0) 


4>1 = -.48 


-3.83 


.000 


Candidate 
Advertising 


(1,1,0) 


4>1 = -.50 


-4.12 


.000 


Contribution 
Limits 


(1,1,0) 


4>1 = -.49 


-3.96 


.000 


Expenditure 
Limits 


(1,1,0) 


<|>1 = -.57 


-4.89 


.000 


Political 
Corruption 


(1,1,0) 


4>1 = -.53 


-4.45 


.000 


Reform 
Legislation 


(0,1,1) 


<|>1 = .58 


4.80 


.000 


Reform 
Proponents 


(0,1,1) 


4>1 =.86 


9.51 


.000 






The ARIMA parameters for each series of the congressional secondary 
sub-issue attributes are depicted in Table 17. Each of the chosen coefficients 
represent well-fitted models for these series (p< 05). 



109 



Table 17: Congress Secondary Sub-Issue Attributes 
ARIMA Analysis of Univariate Series 



Series 


ARIMA 
term 


Coefficients 


t ratio 


P 


Free 
Speech 


(1.1.0) 


<|>1= -.49 


-3.97 


.000 


Soft Money 


(1,1,0) 


<t>1 = -.49 


-3.93 


.000 


Candidate 
Advertising 


(1,1,0) 


<|)1 = -.49 


-3.97 


.000 


Contribution 
Limits 


(1,1.0) 


<|>1 = -.49 


-3.97 


.000 


Expenditure 
Limits 


(1,1,0) 


<|>1 = -.49 


-3.97 


.000 


Political 
Corruption 


(1,1,0) 


<t>1 = -.48 


-3.95 


.000 


Reform 
Legislation 


(1,1,0) 


<|>1 = -.48 


-3.91 


.000 


Reform 
Proponents 


(1,1,0) 


<t>1 = -.49 


-3.97 


.000 



Once the appropriate ARIMA models were determined for each univariate 
series in the two agendas, bivariate relationships were examined using cross- 
correlation analysis. This revealed several significant connections between 
the media and congressional use of secondary level framing attributes. 
These are depicted in Table 18 with lags ranging from -7 to +7 weeks. 



110 



CO 

o 

5 



o 



o 



oo 



o 



eg 
cn 



o 



oo 
o 



o 



o 
o 



-3- 
o 



o 
o 



CO 

o 



+ 



CD 

+ 



CD 



CO 

CD 



C 

E 

CD 
LL 
O 
Q. 
O 

2 

CO 

£> CD 
CD tl 
Xi Q) 
C - 
O 
O -n 

5 c 

CO CD 

o OT 
u CO 

CO CD 

C «- 

2 ? 
In 

CD u 

C 

o 
O 
■o 

CD 
O) 
O) 
TO 
i 

CO 
CO 

o 



CO 
CO 

CD 

I— 

O) 

c 
o 
o 



o 



o 

C 

I 



o 

CO 

n 

CD 

H 



<o 
I 



o 



o 
cn 



in 
o 



en 
oo 



* 

00 

in 



* 
in 



o 



CO 



CN 



00 

o 



CD 



* 

00 

00 



o 



* 

CD 

CO 



* 
CD 






CD 

o 



a> 


m 


o 


r 




>s 




CD 




c 


f 


o 


ree 
peec 


o 


U. CO 


CO 



CN 



CO 



o 



o 



in 
o 



o 



CO 

o 



CO 



oo 
o 



* 
o 

CO 



CO 



00 



00 

o 



CO 



CD ? 
« ._ 
CD (0 

^ CD 

c 5 

CO "D 



q 


q 


q 


CM 

i* 


CN 

q 
■ 


CO 

q 

l 


CN 


CN 

q 
i" 


l" 


q 


a> 
q 


o 

q 

i" 


o 


O) 


CO 



CN 



* 

CN 

CO 



CN 



<N 



c 
g 

B| 

OZ3 



55 
1 



oo 
o 



o 



oo 



oo 



* 



* 



CD 



CM 

o 



o> 



2 

3 



LU _l 



oo 



c 
o 

Si 

52 ■— 
o o 
Q-O 



C» 

q 


i* 


in 
q 


in 
r 


CO 

o 


a> 


i 


i 


in 
o 


in 
o 


■ 


■ 


CN 

q 
i" 


CO 


CO 

q 


T— 


CD 
CN 


q 


* 
CN 


-*• 


* 
in 


CN 


* 

o 

in 


* 
CO 


* 
en 

CN 


* 
CN 


00 

o 


m 

CN 



c 
o 

co 



CD 
01 _J 



■ 

c 

EC 
c o 

O Q. 

■85 2 
a: 0. 



in 

+ 



+ 



CO 

+ 



CN 

+ 



CN 



CO 



in 



CD 



CO 



o 

CD 
Q. 
CO 
CD 



CO 

T3 
CD 

E 

CD 
CO 

■a 

CD 

a) 

CO 
CO 

CD 

I— 

O) 

c 
o 
o 

CO 

r: 

CD 
O 

c 


T3 



CD 
N 

O) 

C 

TJ 

8 

B 

Q. 

CO 

O) 
_CD 

CD 

CO 

C 

g 

CD 

L- 
1— 

o 
o 

Is 

CO a) 

£ .. 3 

O CD J3 

« Z (0 



111 



The results in Table 18 indicate an overall trend for the agenda setting 
effect of the congressional attributes on the media's attributes for 7 out of the 
8 secondary sub-issues examined. The soft money and candidate 
advertising frames reveal the strongest correlations of .67 and .61 
respectively when congressional framing precedes media framing by 4 weeks 
(p<05). Both sub-issues also have significant correlations at weeks when 
the two agendas are considered synchronous. 

The free speech sub-issue frame offers a correlation of .58 when 
congressional framing precedes media framing by 3 weeks (p<05). Many 
other significant relationships are detected in Table 18 with the most common 
direction of influence depicting the congressional framing agenda as a leading 
indicator of the media framing agenda. The time lags range from 2 to 6 
weeks for these correlations. 

Interestingly enough, results also depict a moderate correlation of .40 
when media framing precedes congressional framing by 1 week for the free 
speech sub-issue (p<05). As noted previously, significant correlations were 
found for the reverse agenda setting effect as well (especially at 3 weeks). 
This may indicate a two-way relationship between the two agendas for the 
freedom of speech issue. 

Affective dimension. The second dimension for the framing analysis will 
address the affective attributes used in the media's coverage and 
congressional speeches regarding campaign reform. These will include only 



112 



the negative and positive frames utilized to present the issue for both 
agendas. The neutral frame was not used in Congress, therefore no 
comparison can be offered. 

The ARIMA model parameters chosen for each series of affective 
attributes in the media are depicted in Table 19. Both of the modeled series 
represent significant fits below the .05 level. 





Table 19: Media Affective Attributes 
ARIMA Analysis of Univariate Series 




Series 


ARIMA term 


Coefficients 


t ratio 


P 


Negative 
Positive 


(1,1.0) 
(1.0,1) 


<|>1 = -.44 
<|>1 = .92 &. .83 


-3.47 
6.12 & 3.70 


.001 
.000 & .000 



Table 20 provides the ARIMA parameters chosen for the same two series 
of affective attributes with respect to congressional speeches on campaign 
reform. All models again represent a significant fit. 



Table 20: Congress Affective Attributes 
ARIMA Analysis of Univariate Series 



Series ARIMA term Coefficients t ratio 



Negative (1,1,0) <(>1 = -.49 -4.02 .000 

Positive (1,1,0) f| = -.48 -3.82 .000 



After the models were derived for the affective attributes in both agendas, 
bivariate relationships were examined using the cross-correlation analysis 
technique. Results are presented for this analysis in Table 21 . 



113 











o 


+ 










i 

o 


q 


1 








































2 


o 
i* 


o 
q 


CD 

+ 




CD 
O 








o 
r 


00 

q 


in 

+ 




CD 
Q. 
CO 








o 


CO 


■sj- 




CD 








q 


q 


+ 




£ 


CO 






h- 


CM 


CO 




CD 






q 


q 


+ 




TJ 


3 














CD 


-Q 














E 


< 






CO 


CO 

q 
i 


CN 

+ 




CD 

.C 


c 














CO 
T3 


'e 






CD 


CO 


^_ 




CO 


03 






q 


q 


+ 




_g> 


IX. 




(0 

3 










CO 
CO 


0) 

> 


co 

CO 


o 

c 


co 


o 









U-» 


CD 


6 

-c 

Cj 


q 


T - 


O 




O) 


|.9 












c 
o 


£ -o 


c 


c 










o 


< CD 


o 


is 


o 


(V. 








S^ 


o 


CO 




q 


t — 
i 




"co 


O _ 














.c 


_ T3 














-•^ 


elations 
ress an 






CO 

q 


in 
q 


CN 

i 






o 

c 

CD 














TJ 


Corr 
Cong 






CN 


q 


m 




O 

CD 
N 


CD 






* 


o 


« 




C 


O) 






to 


CN 






TJ 


jo 














8 


CO 














CD 


CO 

o 

l_ 






* 

CJ) 

q 


in 

CN 


ifl 




i_ 
Q. 

CO 


O 














O) 

ro 


T— 

<N 




CO 

% 


q 


w 

CN 


CD 

i 




"co 


(D 




1 








CO 


CO 




8 

1 


(0 

CD 
> 


o 

CN 

.1 


1^- 




c 
g 

CO 
0) 

V— 

o 
o 








*-» 






ii 

11 

CJ).b 








CO 
CD 



z 


o 

0- 


ro 

_i 


















CO <D 
















15 "3 
















° CD X> 








to 








X O ■•-' 








XI 








It Z, CO 








CD 
















2 











114 



The comparison of negative attributes used by the media and congressional 
agenda revealed a relatively strong correlation of .56 when the congressional 
frames precede the media frames by 4 weeks (p< 05). This relationship 
persists at 5 weeks with a correlation of .39 (p< 05). This suggests an 
agenda setting effect of Congress on the media for the use of negative 
attributes. 

The comparison of positive attributes used by the media and Congress 
failed to produce any significant correlations. The correlations averaged a 
low .10 between the two agendas for the positive affective attribute. This 
implies that no agenda setting relationship exists between the two agendas 
for the use of positive attributes. 

Cognitive dimension. The third and final dimension for examination will 
be the cognitive attributes revealed in the media's coverage and 
congressional speeches regarding campaign reform. This will include the 
analysis of causal frames, solution frames, as well as the use of causal and 
solution frames combined. 

The ARIMA parameters for the modeled series of cognitive attributes in 
the media are presented in Table 22. All three coefficients significantly 
represent well-fitted models for these series. 















115 












Table 22: Media Cognitive Attributes 
ARIMA Analysis of Univariate Series 



Series 


ARIMA term 


Coefficients 


t ratio 


P 


Causal 


(0,1,1) 


(j)1 = .70 


6.32 


.000 


Solution 


(1,1,0) 


<)>1 = -.46 


-3.61 


.001 


Causal & 
Solution 


(1,1,0) 


4>1 = -.37 


-2.81 


.007 



Table 23 provides the ARIMA parameters for the congressional series of 
the same cognitive attributes. The analysis of the univariate series indicates 
an appropriate fit for each of the chosen models (p<05). 



Table 23: Congress Cognitive Attributes 
ARIMA Analysis of Univariate Series 



Series 


ARIMA term 


Coefficients 


t ratio 


P 


Causal 


(1,1,0) 


<|>1 = -.50 


-4.06 


.000 


Solution 


(1,1,0) 


<|>1 = -.48 


-3.91 


.000 


Causal & 
Solution 


(1,1,0) 


<M = -.48 


-3.91 


.000 



Once the ARIMA models were determined for both the media and 
congressional series, bivariate associations were examined using cross- 
correlation analysis. Table 24 offers the results for the cross-correlation of 
the cognitive framing attributes. Results indicate numerous significant 
relationships. 



116 



CO 



o 

2 



o 
o 



o 



5 T- 



co 



O) 



00 

o 



+ 



CD 

+ 



lO 

+ 



CD 



o 

CD 
Q. 

co 



CO 





c 

e 

CD 

i_ 

LL 

CD 
> 

o Vt 

O J 



05 

o 



o 



o 
o 



CO 
CO 

c 
o 
O 



% 

o 

c 

8 

i 



co c 

C CD 

2 co 

-™ CD 
O C 

"o 

■D 
CD 
O) 
D) 
JD 
i 

CO 
CO 

o 

I— 

O 

"ST 
CM 

JD 

n 

CD 






O) 



CO 



co 
m 



* 

CN 

CD 



o 



co 

CN 



CO 

tn 

D 
CO 

O 



CN 



O) 
O 



G> 



« 
CN 



O 

CN 



LO 
O 



CD 
CN 



* 
CO 



* 

o> 

CN 



LO 

o 



CO 



c 
o 

o 
co 



m 
o 



in 



co 
o 



+ 



CO 

+ 



CN 

+ 



CD 



o 



* 

v- 
lO 



CO 

o 



CN 
CN 



C 

™ c 
co .2 

CO o 
OCO 



CN 



CO 



LO 



CD 



CO 



CD 

CD 

E 

CD 

.C 

■4-1 

CO 
T3 
CD 
© 

CO 
CO 

2> 

C 

o 
o 

CD 



CD 

CD 
O 

c 

CD 
■D 



CD 
N 

O) 

C 

8 

CD 

i_ 

Q. 

CO 
O) 
_CD 

CD 

CO 

C 

g 

_C0 
CD 

C 

o 
o 

~ T3 
C CD 
CD 



O 



CD 



CD 

CD 



fl 

co a) 

9 o-Q 



z « 



117 



The underlying trend apparent in the results (Table 24) suggest that 

congressional framing has an agenda setting effect on media framing for all of 

the cognitive attributes. The strongest correlations are detected when the 

congressional frames precede the media frames by 3 to 6 weeks. A very 

large correlation of .70 appears specifically for the use of combined causal 

and solution attributes when the congressional agenda precedes the media 

agenda by 4 weeks (p<05). The agendas also have a synchronous 

correlation of .42 for the use of solution framing attributes at a week lag 
(P<05). 

Research question 7 specifically addressed the relationship between the 
attributes used in the media agenda and the attributes used in the 
congressional agenda for the campaign reform issue. This analysis offered a 
three dimensional examination of the attributes used for both agendas. 
Primary sub-issues, secondary sub-issues, affective attributes, and cognitive 
attributes were used to provide a complete picture of the framing effects. 
Results overwhelmingly indicated an agenda setting influence of the 
congressional frames on the media frames for the campaign reform issue. 
This held true for each of the framing dimensions analyzed. Further 
discussion of these and previous results will be addressed in the following 
chapter. The final chapter will also detail the limitations of this study and 
suggestions for future research. 



CHAPTER 5 
DISCUSSION 

Traditional agenda setting has for years asserted that the salience 
afforded an issue in one agenda will affect the salience of that same issue 
with respect to other agendas (McCombs & Shaw, 1972). This is a time- 
related process that has been studied with a variety of methodologies. Based 
on this accumulation of research, the concept of the agenda setting process 
has often addressed the following three components: the media agenda, the 
public agenda, and the policy agenda (Rogers & Dearing, 1988; Gonzenbach, 
1992, 1996). 

Recent efforts have expanded the agenda setting theoretical framework to 
include a second level (McCombs, 1981, 1992; McCombs & Evatt, 1995; 
Ghanem, 1996, 1997; McCombs & Ghanem, 1998). Agenda setting is no 
longer limited to the study of issue salience. As McCombs and Evatt (1995) 
state, "in addition to providing cues about the salience of topics-objects, if you 
will- the mass media differentiate between the saliency of various attributes of 
these topics or objects" (p. 134). In this respect, what (objects) as well as how 
(attributes) an issue is emphasized may have an agenda setting influence. 

This study was specifically designed to examine the traditional first level of 
agenda setting as well as the second level of agenda setting for the campaign 
reform issue in 1999. The importance of further research bridging the two 

118 



119 



levels of agenda setting is vital in the development of an expanded 
theoretical paradigm. The present analysis included representations of the 
media agenda, the public agenda, and the congressional agenda. Articles 
from the New York Times and the Washington Post were used to depict the 
media agenda. Speeches on the floor of the House and Senate portrayed the 
congressional agenda. Finally, messages from USENET on-line discussion 
groups represented the public agenda for the issue. 

The examination began by scrutinizing the salience of the campaign 
reform issue as an object on the three agendas. This involved frequency 
counts for each agenda to represent the amount of attention afforded the 
issue. This was then expanded to include an analysis of the attributes for the 
issue as well. An analysis of the possible bi-level effects between issue 
salience and issue attributes was also included. The overall objective of the 
investigation was to capture a complete and accurate picture of the forces 
that sculpt the three agendas. This incorporates both how much they think 
about the campaign reform issue and then specifically what they think 
regarding the campaign reform issue. 

The study analyzed the campaign reform issue as it appeared in the 
media, public, and congressional agendas for the 52 weeks of 1999. Items 
were coded for the three agendas and cross-correlations were established 
using ARIMA time series models. First level agenda setting was represented 
by the frequency of emphasis, while second level agenda setting was 






120 



represented by four framing dimensions. The four dimensions have been 
utilized in past research (Ghanem, 1996, 1997) and include the following: 
subtopics, framing mechanisms, affective attributes, and cognitive attributes. 
Framing mechanisms were addressed at the descriptive level, while the 
remaining three dimensions were used to provide an in-depth time series 
analysis of the attributes. 

Results for the first level analysis suggest that the most immediate 
significant agenda setting effect is between the public and media agendas. 
They are correlated in their emphasis of the campaign reform issue at week 
(synchronous) and when public attention on the Internet precedes media 
coverage by 1 week. Past research has revealed that media coverage often 
reflects consumer desire (Graber, 1979; Gonzenbach, 1992). A summation 
of all significant first level results is depicted in Figure 17. 



Public 

Congress 

Public 




Congress 



3 4 

Weeks 



Figure 17: First Level Agenda Setting Relationships for the Media, Public, and 
Congressional Agendas (• p<05, For example: Line three denotes that the public 
leads Congress with .05 significance at a 2 and 3 week lag) 



The next first level agenda setting effect appears to be of the public 
agenda on the congressional agenda at a lag of 2 to 3 weeks. This too has 



121 



been found in past research (Page & Shapiro, 1983; Converse, 1987; 
Gonzenbach, 1992). It is commendable to think that the public should exert 
such an influence in shaping congressional debate. The use of Internet 
discussion surely represents a more politically active public and appears to 
exert a powerful effect on those in policy positions. 

A significant agenda setting effect is then noted when congressional 
attention precedes media attention by 4 to 6 weeks. This is not the first study 
to reveal that policy elites may affect media coverage (Walker, 1977; Behr & 
Iyengar, 1985; Rogers, Dearing, & Chang, 1991). 

Finally, the public agenda is again correlated with the media agenda at a 
lag of 3 to 7 weeks. This may indicate that overall the public begins by 
voicing their concerns about campaign reform on the Internet, the media then 
quickly picks up on them with increased coverage. Congressmen also heed 
these public concerns and in turn speak about them 2 to 3 weeks later in 
Congress. The media cover the issue in response to congressional debate 
approximately 4 to 6 weeks later. At this time the public and media agendas 
become highly correlated (Figure 17). In this respect, we can be optimistic in 
stating that in many ways the public shapes the concern for the issue of 
campaign reform. 

Results for the second level analysis begin by offering a bi-level 
investigation of the agenda setting relationship between the salience of 
campaign reform on the public agenda (first level) and the attributes used 



122 



regarding the issue in the media (second level). Figure 18 offers a portrait 
summarizing these relationships for the campaign reform issue. 



Public 
Public 
Public 














Media 

Sub-Issues 

Media 

Cognitive 
























Affective 















1 


2 


3 4 
Weeks 


5 


6 7 





Figure 18: Bi-Level Agenda Setting Relationship between the Public Agenda and 
Media Framing Attributes (• p<.05, For example: Line two denotes that the public 
leads media affective attributes with .05 significance at a and 5 week lag) 



Results indicate that the public agenda is a leading indicator for the 
frequency of framing attributes utilized by the media. This holds true 
especially for the media's use of subtopics and cognitive attributes, but is 
weaker with respect to the use of affective attributes. It is plausible that in this 
instance the tone of media coverage for these two newspapers is at times 
more likely to be independent. These relationships are surprising because 
past research suggests the opposite, that the number of attributes in the 
media compels the public to assign increased salience to issues (McCombs, 
1996; Ghanem, 1996, 1997). Perhaps, the reverse is true in this case 
because the public agenda is represented by those taking the more 
aggressive step in communicating their concerns on the Internet. These 



123 



individuals are more likely to be politically active and their voice appears to 
exert a powerful effect. 

The second level of agenda setting is again addressed with an 
examination of the media and congressional attributes used to frame the 
campaign reform issue. This final investigation differs from the previous in 
that it offers a comparison that is strictly concerned with the attributes utilized 
by each agenda. Figure 19 presents a summary of the significant 
relationships apparent in the second level agenda setting for media and 
congressional attributes. 

Several attributes are displayed indicating a variety of relationships 
between the two agendas. The obvious trend suggests that the 
congressional framing agenda leads the media framing agenda by an 
average of 3 to 6 weeks. Past research has also indicated such a 
relationship between the policy and media agendas (Entman & Rojecki, 1993; 
Reese & Buckalew, 1995). This correlation held true for the majority of 
attributes, with only two (positive affective and expenditure limits) failing to 
produce a significant relationship. The first is understandable considering 
that the positive tone is arguably the least attractive for media coverage. A 
lack of significant results regarding expenditure limits can be attributed to the 
fact that this sub-issue simply received minimal attention from either of the 
two agendas (Figure 19). 



124 



Media 

Free 
Speech 

Reform 
Legislation 

Congress 

Free 
Speech 

Soft 
Money 

Candidate 
Advertising 

Contribution 
Limits 

Political 
Corruption 

Reform 
Legislation 

Reform 
Proponents 

Negative 
Affective 

Causal 
Cognitive 

Solution 
Cognitive 

Causal & 

Solution 

Cognitive 




Congress 
Congress 



3 4 

Weeks 



Media 



Media 
Media 




Media 



Media 

Media 
Media 



Media 
Media 

Media 



Figure 19: Second Level Agenda Setting Relationships for the Media and 
Congressional Framing Attributes (• p<05, For example: Line one denotes that the 
media leads Congress for the free speech frame with .05 significance at a 1 week 
lag) 



Two of the framing attributes (free speech and reform legislation) also 
suggested an agenda setting effect when the media precede Congress by 1 
week. This may indicate a two-way relationship for these sub-issues. The 
media preceding in the emphasis of the reform legislation frame can be 
explained as anticipatory coverage of scheduled reform debate in Congress. 



125 

The media's influence with respect to the free speech frame is more difficult 
to interpret. It may simply be a result of the fact that free speech is a sub- 
issue that is inherently near and dear to the hearts of journalists. 

Overall, the findings in this study indicate that the agenda setting process 
for the campaign reform issue is very complex. A variety of significant effects 
were noted, but some underlying trends did emerge that lend further insight 
into agenda setting theory. It would appear that for the first level of agenda 
setting, public concern regarding campaign reform propels the issue. This 
influence may be a direct result of the representation of the public agenda by 
those taking the more active step in discussing their opinions on the Internet. 

Results suggest that such an influence persists at the second level of 
agenda setting, at least with respect to the public agenda leading the media in 
their choice of attributes. Purely second level analysis of the public influence 
has yet to be determined, but once the framing attributes are discussed by 
Congress they have a powerful effect on the attributes utilized by the media. 

The agenda setting paradigm can benefit from these results because they 
offer yet another perspective in the development of a multi-dimensional 
agenda setting theory. The public is a powerful force on the Internet and this 
has an affect on the study of traditional as well as second level agenda 
setting. These findings are surprising in some ways, but refreshing in an age 
where many perceive the media to be an inappropriately powerful leader in 
political discourse. 



126 



The limitations of this study are predominantly directed at the lack of 
access to public discourse regarding the campaign reform issue. The choice 
of on-line discussion messages to represent the public's framing of the issue 
proved to be problematic. The unavoidable loss of archives due to the 
database changing hands was beyond the researcher's control, yet it left a 
hole in the examination of second level agenda setting effects. Future study 
must incorporate an element of public discourse to make the analysis of this 
issue complete. 

The use of on-line sources to represent the public agenda at any level may 
be questionable. The databases have yet to be perfected and often produce 
samples that may not adequately display true public concern with an issue. 
The Internet represents a more politically active public and that must be 
considered when analyzing results. The Internet is a new frontier for 
researchers, but must be utilized with care. Future analysis of Internet 
sources should strive for maximal control over the research situation. 

The current study examined the issue of campaign reform for a one year 
time period. This in itself can be viewed as a limitation. Many longitudinal 
studies cover a span of at least 3 to 6 years in an effort to display the ebb and 
flow of agenda setting relationships. Future research may be needed to 
expand upon the present findings with the addition of multiple years of 
analysis. It would be interesting to note the possible changes in agenda 
setting influence with respect to election year debate on campaign reform. A 



127 



major candidate in the 2000 presidential election maintained campaign reform 
as his dominant platform. This persistent focus reveals a concern with the 
issue of campaign reform that has become a significant part of the American 
political climate in recent years. The Senate has currently (April 2001) 
passed a campaign reform bill that will reach the House of Representatives 
for debate in July 2001 . These events have surely had an impact on the 
issue and the attention afforded it. A study is needed that can capture more 
of the Senate discussion regarding the issue. This study failed to do so 
because of the speech format chosen for analysis. 

This analysis focused entirely on one issue in the political climate in order 
to gain an accurate depiction of that issue on the three agendas. Future 
study may wish to integrate the examination of multiple issues in order to 
observe the competition and comparison of each with respect to first and 
second level agenda setting effects. An analysis of the "issue attention 
cycles" apparent in the competition among these issues would give insight 
into their intense struggle for survival on the different American agendas. The 
study of a variety of issue types is also warranted in order to develop a more 
accurate theoretical model. 

Limitations often arise with the choices made to portray specific agendas. 
The results can only be applied with accuracy to the sample chosen to 
represent each agenda. The current study is no exception. It is arguable, for 
example, that coverage in the New York Times and the Washington Post 



128 



differs from that in regional publications. We make justifications for these 
choices, but ultimately each source of analysis varies to some degree. For 
this reason, future study must continue to investigate new ways to represent 
each agenda. With each new study, we get closer to understanding the 
underlying dynamic of agenda setting. 

The exploration into second level agenda setting is still relatively new to 
mass communication research. This is a limiting factor in that former 
successful techniques are not readily available for replication. The current 
study looked to past research for the development of a multi-dimensional 
framing approach. These dimensions, while used before, require highly 
subjective choices with respect to each new issue. Future directions toward 
the development of second level agenda setting theory must provide a 
framework within which to make these choices. The agenda setting of 
attributes is an area of great intrigue, but with it comes a great need for 
internal and external consistency. 

The power to influence the salience afforded an issue as well as the 
specific attributes used to discuss that issue is a highly coveted authority. 
The political arena is especially replete with those who are hopeful to assert 
their influence with compelling results. This study went to the heart of political 
campaign reform by examining both first and second level agenda setting 
effects for the issue. Results suggested that in 1999 the public painted the 
initial portrait of campaign reform and the image was recognized in earnest by 



129 



those in Congress and those in the media. Congress quickly picked up the 
paintbrush and with detailed strokes affected the specific images relayed to 
the media. These results both align and depart from past agenda setting 
research. They are above all enlightening and will hopefully generate 
numerous threads of inquiry adding yet another layer to the fabric of agenda 
setting. 






APPENDIX A 
SAMPLE ARTICLE FROM THE NEW YORK TIMES 

4 of 143 DOCUMENTS 
Copyright 1999 The New York Times Company 
The New York Times 



December 17, 1999, Friday, Late Edition - Final 
SECTION: Section A; Page 30; Column 1; Editorial Desk 

LENGTH: 576 words 

HEADLINE: The McCain-Bradley Show 

BODY: 

The air and even some of the questions in New Hampshire were frosty yesterday 
at the unusual joint appearance of Senator John McCain of Arizona and former 
Senator Bill Bradley of New Jersey. But these two presidential candidates were 
making a significant contribution with their attempt to give the issue of 
campaign finance reform a more prominent place on the agenda for 2000. This 
marked the first time that two serious contenders from opposing parties have 
come together to challenge the Washington establishment on so fundamental a 
problem as the two parties' joint dependence on special-interest money. 



130 



131 



It was especially fitting that Mr. McCain and Mr. Bradley made then- 
appearance, broadcast on ABC's "Nightline" last night, at the exact spot where 
President Clinton and former House Speaker Newt Gingrich promised to bring about 
a bipartisan campaign cleanup in 1995. Nothing came of that meeting except for 
mounting popular disenchantment as detail after detail of the corrupt election 
of 1996 came to light. It is still fashionable in Washington to think that 
campaign finance reform is a low priority among voters, as Ted Koppel said 
repeatedly. But in fact, the issue has propelled both Mr. McCain and Mr. Bradley 
in New Hampshire, where they are surging ahead because they have made it a 
centerpiece of their respective campaigns. 

As expected, Mr. McCain and Mr. Bradley exchanged promises to bar their 
parties from spending the unlimited contributions by corporations, unions and 
rich individuals known as soft money should they face one another in the general 
election. These donations to the party circumvent longtime limits on campaign 
donations to the candidates enacted by Congress over many decades, and Mr. 
McCain said he would forgo soft money no matter whom he faces. Vice President Al 
Gore has also pledged to forgo soft money in a general election if his opponent 
does. But Mr. McCain and Mr. Bradley have been most aggressive in promoting a 
move that would cut off an expected $500 million in party donations from the 
nation's most powerful special interests. 

Significantly, yesterday's joint appearance by Mr. McCain and Mr. Bradley 
drew an irate response from George W. Bush. Early in the campaign the Texas 
governor said he would favor a ban on soft-money donations to political parties 
by corporations and unions. Lately he has retreated, saying that such a ban 
would be effective only if it were attached to a harshly anti-labor provision 
making it difficult for unions to raise campaign money from their members. His 
position is a dodge, since corporations already give more to campaigns than 
unions do, by a ratio of more than 10 to 1. 

It took some nerve for Mr. McCain to attack his party's front-runner on this 
issue. Mr. Bush has rejected fund-raising limits and is rapidly becoming the 
champion money harvester of all time, even compared with Bill Clinton By 
defending soft money, Mr. Bush, a self-styled outsider, has truly become the 
advocate of the Washington fund-raising establishment in this election. 

In hard political terms, Mr. Bradley and Mr. McCain were competing yesterday 
for the independents and crossover voters that both of them need in New 
Hampshire's open primary. But their meeting has the potential of expanding the 
base of voters who understand the corrupting influence of special-interest 
money. They should take their show on the road and invite other candidates to 
join them. 

http://www.nytimes.com 



APPENDIX B 

SAMPLE "MOST IMPORTANT PROBLEM" POLLS FOR THE 

CAMPAIGN REFORM ISSUE 



Poll Questions Regarding Campaign Reform (CR) for 1999 


11/1994 


Exit poll: Does campaign reform matter in 
your choice of a candidate? 


Yes- 96% 


11/1994 


Thinking about congress, do you think 
(campaign finance reform) should be high, 
low, or not a priority? 


High- 45% 
Low- 36% 
Not a prior- 12% 
Not sure- 7% 


8/1995 


Who do you feel would do a better job 
dealing with campaign finance reform? 


Rep.- 26% 
Dem.-26% 


10/1995 


What one issue would you like to hear pres. 
candidates talk more about? 


CR- 4% 


5/1996 


Who do you feel would do a better job 
dealing with campaign finance reform? 


Dem.- 20% 
Rep.- 19% 
Both- 18% 


7/1996 


I am going to read a list of issues facing the 
nation... To what degree is (campaign 
finance reform) an important issue? 


Extremely- 12% 
Very Imp- 29% 
Somewhat- 35% 


9/1996 


What one issue would you like to hear pres. 
candidates talk more about? 


CR- 3% 


10/1996 


Which of the following things would you like 
to see most in the next election? 


CR- 25% 
Less negative 
campaigning- 35% 
Higher voter turnout- 
23% 


10/1996 


What do you think is the most important 
problem for the president to deal with... 


CR-1% 


11/1996 


I am going to read you a list of issues. Tell 
me which should be the highest priority for 
the president?... 


CR-1% 


11/1996 


What do you think should be the top priority 
for the next administration? 


CR-1% 








1/1997 


What do you think are the two most 
important issues that the federal 
government should address? 


CR-1% 



132 



133 



2/1997 


I am going to read you a list of issues. Tell 
me which should be the highest priority for 
the president?... 


CR- 3% 


3/1997 


Do you think that you are more likely to see 
Elvis or to see real campaign finance 
reform? 


See Elvis- 48% 
SeeCR-31% 


4/1997 


What do you think is the most important 
problem facing congress? 


CR- 2% 


4/1997 


I am going to read you a list of issues. Tell 
me how much of a priority (campaign 
finance reform) should be? 


Top- 11% 
High- 33% 
Low- 44% 


7/1997 


What do you think is the most important 
problem facing congress? 


CR-1% 


8/1997 


Who do you feel would do a better job 
dealing with campaign finance reform? 


Dem.- 24% 
Rep- 36% 


9/1997 


Regarding political action com: Should the 
system be changed to reduce the role of big 
money? 


Yes- 78% 
No- 15% 


9/1997 


Which of the following is closer to your point 
of view? CR is needed cause too much $ in 
politics... or ...CR not needed b/c $ 
contributions should be protected by 
freedom of speech 


Needed- 77% 
Not needed- 18% 


10/1997 


Do you agree or disagree that money is 
equal to free speech? 


Agree- 50% 
Disagree- 37% 
Not sure- 13% 


10/1997 


Would you support a candidate more or less 
if they voted for campaign finance reform? 


More- 31% 
Less- 4% 


10/1997 


Who do you feel is to blame for the failure to 
pass campaign reform legislation this year... 


Dem- 22% 
Rep.- 28% 
Both- 35% 








10/1997 


I am going to read you a list of issues. Tell 
me how much of a priority (campaign 
finance reform) should be? 


Top- 16% 
High- 32% 
Low- 39% 


10/1997 


Who do you feel would do a better job 
dealing with campaign finance reform'? 


Rep.- 37% 
Dem- 31% 


10/1997 


How strongly do you feel there is a need for 
campaign finance reform? 


Strong- 44% 
Somewhat- 24% 


1/1998 


As pres. Clinton prepares to give the state 
of the union address, there are many issues 
facing the nation. I will read a list and tell me 
if you think it should be a priority for 1 999 


CR 

High priority- 45% 
Low- 52% 
Not sure- 3% 



134 



4/18/98 



6/24/98 



6/24/98 



6/24/98 



6/30/98 



8/6/98 



9/10/98 



10/28/98 



11/5/98 



1/1999 



1/1999 



2/24/99 



4/5/99 



5/19/99 



Who do you feel is to blame for the failure to 
pass campaign reform legislation this year... 



Who do you feel would do a better job 
dealing with campaign finance reform? 






Which of the following reform do you feel is 
the most important to address? Fed. tax, 
healthcare, campaign finance, social sec... 



Rep.- 31% 
Dem-18% 



Rep.- 32% 
Dem.- 26% 
Both- 7% 
Neither- 15% 
Don't know- 19% 



CR- 7% 



Which is the second most important? 



Thinking about congress, do you think 
(campaign finance reform) should be high, 
low, or not a priority? 



Read a list of issues... Would (campaign 
finance reform) be an important factor in the 
way you vote? 



If a candidate voted against campaign 
finance reform, would you still support that 
candidate? 



What do you think are the two most 
important issues that the federal 
government should address? 



CR-10% 



High priority- 38% 
Low priority- 41 % 



Most Imp- 15% 
One of many things- 
45% 



Yes- 10% 



am going to read you a list of issues. Tell 
me how much of a priority (campaign 
finance reform) should be? 



Which of the following issues should receive 
highest priority from congress... economy, 
campaign finance reform, taxes, moral 
issues 



What do you think are the two most 
important issues that the federal 
government should address? 



CR- 2% 



Top- 50% 
Lower- 34% 
Not at all- 11% 
Don't know- 5% 



CR- 3% 



What do you think are the two most 
important issues that the federal 
government should address? 



Which of the following issues should get the 
highest priority from congress.. 



What do you think are the two most 
important issues that the federal 
government should address? 



CR-1% 



CR- 2% 



CR- 2% 



CR-1% 



135 



6/4/99 


I am going to read a list of issues, tell me if 
you think a candidate's position on that 
issue would play a role in your decision to 
support or not support... 


Top priority- 1 0% 
High prior.- 14% 
Very Imp- 29% 
Somewhat- 26% 
Not Imp.- 18% 
No opinion- 3% 


6/29/99 


If congress were to pass campaign finance 
reform legislation, do you think it would 
make it easier for challengers to defeat 
incumbents or harder...? 


Easier- 43% 
Harder- 22% 
No diff.- 20% 
Don'tknow-15% 


7/13/99 


What one issue would you like to hear the 
presidential candidates talk more about next 
year? 


CR-1% 


9/8/99 


What do you think are the two most 
important issues for our federal government 
to address? 


CR- 1 % 


9/14/99 


If you could recommend one change in our 
political system, what would it be? 


CR-11%?? 


10/23/99 


Suppose congress passes campaign 
finance reform legislation. Do you think their 
would be a positive effect, no effect, or don't 
know. . . 


Positive- 43% 
None- 44% 
Don't Know 


11/4/99 


Some people feel so strongly about an issue 
that they will vote for or against a candidate 
solely for their stance on that issue. I'll read 
a list of issues, tell me if you consider any 
so important that you would vote for or 
against a candidate solely b/c of stance... . 


CR- 5% 


12/6/99 


What do you think are the two most 
important issues for the federal government 
to address? 


CR-1% 


12/6/99 


Regardless of which issues you just named, 
which of the following issues do you think 
will be most important to you personally as 
you make up your mind who to support for 
the 2000 election?... healthcare, education, 
social sec, taxes, moral values, economy, 
foreign policy, environment, or campaign 
finance reform? 


CR- 2% 


12/8/99 


Candidates are discusses many issues for 
the upcoming election. I am going to read a 
list and tell me which you have heard 
about? 


CR 

Heard of- 61% 
Haven't- 39% 



136 



12/8/99 


Which candidate do you most associate with 
the campaign reform issue? 




12/12/99 


What is the single most important issue to 
you in deciding who to support for 
prez... education, economy, taxes, social 
security, campaign reform, world affairs, 
healthcare, or others? 


CR-1% 


12/12/99 


Which one of these candidates do you most 
trust to handle the issue of campaign 
reform? 


Bush- 34% 
Gore- 24% 
McCain- 19% 
Bradley- 14% 



APPENDIX C 
SAMPLE MESSAGE FROM DEJA.COM SEARCH ENGINE 

NOTICE: Deja.com will be offline from 6:30 am - 2:30 pm Central Time on Saturday, February 19 
while we upgrade our site. We apologize for the inconvenience. 







Help | Feedback | My Deja 



Home 



rn tilths discussions communities 



Power Search 



f /» . ■• itinswl 






Tt». " * 



Jaarcfr 






F orum : alt. politics 

•ad : Is Senator McCain a Hypocrite About 
Campaign Finance Reform? 
Message 3 of 1486 

Subject: Is Senator McCain a Hypocrite About 
Campaign Finance Reform? 

Date: 02/07/2000 

Author: f'om Abbott <tabboll .'<ji»Uejngxxotn> 

05 Save this thread 

The Bush supporters have another theme they are using to try and damage 

Senator McCain's Presidential campaign. Some members of the press have 

also followed this same twisted logic. 

The argument is that if Senator McCain is promoting campaign Finance 
reform, then he should not accept ANY campaign contributions (or at 
least this is the implication they leave). Since McCain DOES accept 
campaign contributions, McCain's opponents say he is being 
hypocritical for accepting money while calling for campaing finance reform. 

What these people are missing, either inadvertenly or deliberately, is that 
McCain's finance reform policy does not call for the elimination of ALL 
campaign contributions, it just calls for more control on existing 
contributions, and the elimination of some types of 
contributions like "soft money." 



baokto searcliresults 






Auction Showcase 
Find It at: --J, 

o 

*- vi mid Crafls S j 

♦ i; 1 1 n -i jgl 

( ?;■■!■ :• ies 
♦■ Dail\ I .- M | i ■ , 

' 3 f Rf P 




I 

Search for Jobs! 
■JobOntions 

Seen it on TV, in 

print? 

Find (lie book <u> 

AMAZON.COM 

VHl.com 
Valentine Day 
Send a cool 
MFSlCcanl 



137 



138 



McCain's campaign finance reform would allow each voter to contribute 
$1,000 to the candidate of his or her choice, and that would go for those 
working for corporations and big business, too. All soft money would be 
eliminated, which is where the opportunity for government corruption is so 
much greater because of the amounts of money 
involved 

So McCain's reforms would level the playing field between the 
average citizen and the tat cat by restricting both to contributions of $1,000 
each The millions of dollars of soft money the fatcats use to gain an 
advantage over the private citizen in the "influence arena," would be 
eliminated And should be. 

So, McCain is not being a hypocrite. He is conforming to the campaign 
finance rules that he is proposing that everyone should follow. He only 
accepts $1,000 dollar donations, and that goes for the fatcats and lobbyist 
who support him, too. 



Great Music? 
Click Here! 

FREE 100 min. 
Phone Card! 
When you <ry 
Nctmarkct. 

Deja.com is 
Hiring!! 
(lick here for 
jobs 



TA 

1'osl Kctth 



I'm butt in h anil • >.i .\i in search 



O Save this thread 
n! .cn.bvto alt.politics 

Vjjiijj 1 1 p • - i oc >>•;;..;.• !_0 • ! fi Ij.'Kl 
\ sev. ?riii»ia! I >tt>ct foi mat 
Crvote. ■'• cuviom link to th»> inv->b<<£e from \vu\ cmn \\eb >it c 

For a more detailed search in Discussions go to l'ou.i Search 
Search onU in: ' ' alt.politics 

'• All Deja.com 
* earcn ' or: ! campaign reform or campaign finance rcfonr j Search j 



Search recent 



Messages 



Arts & t.iiicrtainiiicn t Automotive Commnini: j£ lech Consumer Electronics Health g 

Fitness 
Home cS, Fa mily Lilcsti \vs .Money Politics & Media Recreation Spoils Traycl 

C«.»U'.pulW"s [Sure '"'i 1 - Get Free SjulTaJ Fre eSliopi - ComputerJobs - 

Gel I HI I Ik-aUi InkvwdiL.iop uhu - [iteRack com - BJ iY AT COST at eCOST.com - 
Search l^i lobv JobiJpjion? - Sjearchjjsed cars - Onsaie Auctions - VHI c om - Sonic ne t 

I opvriuhj © 1995-2000 
Ueia c«wh, Inc All rights 
reserved. 

I I adciiiaii •. • I enns & 

! r.MK!j.ti<*!Vj ofCstf • Sl!C 

l';r uc\ Stalunenl. 



APPENDIX D 
SAMPLE SPEECH FROM CONGRESSIONAL RECORD 



CONGRESSIONAL RECORD - HOUSE 

Tuesday, September 28, 1999 

106th Congress, 1st Session 
145 Cong RecH 8877 
REFERENCE: Vol. 145, No. 128 

TITLE: ENFORCE EXISTING CAMPAIGN FINANCE REFORM LAWS 
SPEAKER: Mr. PITTS 
TEXT: [*H8877] 

(Mr. PITTS asked and was given permission to address the House for 1 
minute and to revise and extend his remarks.) 

Mr. PITTS. Mr. Speaker, we have been hearing a lot lately about 
campaign finance reform. However, some of those pushing for new laws 
fail to mention the fact that our existing laws have been broken. 

It is against the law to use foreign money in election campaigns in 
America. It is against the law to launder money in election campaigns. 
It is against the law to sell access to your office or influence or 
even seats on a foreign trade mission to highest bidders. It is against 
the law to use public offices, telephones, equipment, staff, computers 
in election campaigns. 



139 



140 



We have heard about "no controlling legal authority." The Attorney 
General not only fails to enforce our existing laws on campaign finance 
reform, but the Attorney General blocks efforts to investigate existing 
laws. 

We should have full disclosure, but we should also have our existing 
laws enforced It is a scam on the American people to pass new laws on 
finance reform, while not enforcing existing laws. 



SUBJECT: CAMPAIGN FINANCE (93%); CAMPAIGNS & 

ELECTIONS (90%); LEGISLATION (90%); MONEY LAUNDERING (55%); 

LOAD-DATE: 09-29-1 



APPENDIX E 
CODING SHEETS FOR FRAMING CATEGORIES 



Content Analysis of Attributes 
Code Book for Media Agenda 



GENERAL INFORMATION 



(filled out by researcher) 
Coded by Date coded 



Source: 

<> New York Times <> Washington Post 

Text Identification # Week of text 



SUB-ISSUE ATTRIBUTES 



Choose and circle one of the following sub-issues to represent the primary 
attribute in the text: 

<> Freedom of speech: any discussion of money as speech or 1 st 

amendment rights 
<> Soft Money: discussion regarding unregulated contributions from 

special interest groups (unions, corporations, etc.) 
<> Candidate advertising: reference to political advertisements, free 

commercial time, express advocacy issues 
<> Incumbency Advantage: reference to incumbent resources and 

increased access- for example: franking privileges, gerrymandering 
Contribution Limits: discussion regarding limiting campaign 

contributions & incentives to do so 



141 



142 



Expenditure Limits: discussion regarding limiting a candidate's 

campaign expenditures 
Political corruption: reference to American cynicism with the 

democratic process 
Legislation: discussion regarding specific campaign reform 

legislation or bills before congress- for example: 1971 FECA 
Reform proponents: reference to specific individuals such as McCain, 

Feingold, Shays, Meehan 
Other 



Do any of the following sub-issues represent secondary attributes in the 
text?(choose as many as apply, but DO NOT choose your primary attribute): 

<> Freedom of speech: money as speech, 1 st amendment right 
<> Soft Money: unregulated contributions from special interest 

groups (unions, corporations, etc.) 

Candidate advertising: advertising, express advocacy, free airtime 

Incumbency Advantage: due to franking privileges, 

gerrymandering, etc. 

Contribution Limits: regarding limiting amount of contributions, 

incentives 

Expenditure Limits: regarding limiting campaign expenditure 

Political corruption: threat to democracy, pervasive American 

cynicism 

Legislation: example-1971 FECA 

Reform proponents: McCain, Feingold, Shays, Meehan 

Other 









143 



FRAMING MECHANISMS 



(filled out by researcher) 
# of words in the article 

Choose one of the following to represent the location of the article: 

<1> Front main page <2> Front page/inside section <3> Elsewhere in 

newspaper 

Check any illustrative elements used in the article: 

<1> Photographs <2> Graphics <3> None 



AFFECTIVE ATTRIBUTES 



Choose and circle one of the following to represent the tone of the text 
analyzed: 

<> Negative: generally engaging in a negative or judgmental tone, hostile 
or conflictual, using attacks or aggressive language, pessimistic or 
sarcastic evaluations of the topic and those involved (word examples: 
corrupt, disgust, frustration, chaotic) 

Positive: generally making judgments in a positive or optimistic tone, 
stressing fairness and acceptance, presenting views using logical and 
non-inflammatory language, calm and sensitive demeanor (word 
examples: agree, change, compromise, helpful, benefit) 
Neutral: The absence of judgment; statements of fact; nonevaluative 
reports of related events, gatherings, or meetings. 
None of the above 



144 



COGNITIVE ATTRIBUTES 



Do any of the following represent the cognitive structure of the article? (you 
may choose as many as apply): 



Cause: The article refers to attributions of causal responsibility, 

identification of who/what is responsible for problematic campaign 

financing or the absence of meaningful reform. 
Solution: The article mentions the solution to the problem of 

campaign financing and identifies who/what will remedy the 

problem. 
None of the above 



CODER COMMENTS/ IF ANY: 



Content Analysis of Attributes 
Code Book for Public Agenda 



GENERAL INFORMATION 



(filled out by researcher) 
Coded by Date coded 



Source: 

<> Internet discussion 



Text Identification # Week of text 



145 



SUB-ISSUE ATTRIBUTES 



Choose and circle one of the following sub-issues to represent the primary 
attribute in the text: 

Freedom of speech: any discussion of money as speech or 1 st 

amendment rights 
Soft Money: discussion regarding unregulated contributions from 

special interest groups (unions, corporations, etc.) 
Candidate advertising: reference to political advertisements, free 

commercial time, express advocacy issues 

Incumbency Advantage: reference to incumbent resources and 

increased access- for example: franking privileges, gerrymandering 
Contribution Limits: discussion regarding limiting campaign 

contributions & incentives to do so 

Expenditure Limits: discussion regarding limiting a candidate's 

campaign expenditures 
Political corruption: reference to American cynicism with the 

democratic process 
Legislation: discussion regarding specific campaign reform 

legislation or bills before congress- for example: 1971 FECA 
Reform proponents: reference to specific individuals such as McCain, 

Feingold, Shays, Meehan 
Other 

Do any of the following sub-issues represent secondary attributes in the text?( 
choose as many as apply, but DO NOT choose your primary attribute): 

<> Freedom of speech: money as speech, 1 st amendment right 
<> Soft Money: unregulated contributions from special interest 
groups (unions, corporations, etc.) 

■ ■ ■ 

<> Candidate advertising: advertising, express advocacy, free airtime 



<> 



146 



<> Incumbency Advantage: due to franking privileges, 

gerrymandering, etc. 
<> Contribution Limits: regarding limiting amount of contributions, 

incentives 
<> Expenditure Limits: regarding limiting campaign expenditures 
<> Political corruption: threat to democracy, pervasive American 

cynicism 
<> Legislation: example-1971 FECA 
<> Reform proponents: McCain, Feingold, Shays, Meehan 

Other 



FRAMING MECHANISMS 



(filled out by researcher) 
# of words in the message 



# of quote references 



AFFECTIVE ATTRIBUTES 



Choose and circle one of the following to represent the tone of the text 
analyzed: 

<> Negative: generally engaging in a negative or judgmental tone, hostile 
or conflictual, using attacks or aggressive language, pessimistic or 
sarcastic evaluations of the topic and those involved (word examples: 
corrupt, disgust, frustration, chaotic) 

Positive: generally making judgments in a positive or optimistic tone, 
stressing fairness and acceptance, presenting views using logical and 
non-inflammatory language, calm and sensitive demeanor (word 
examples: agree, change, compromise, helpful, benefit) 



147 



Neutral: The absence of judgment; statements of fact; nonevaluative 
reports of related events, gatherings, or meetings. 
None of the above 



COGNITIVE ATTRIBUTES 



Do any of the following represent the cognitive structure of the message (you 
may choose as many as apply): 



Cause: The article refers to attributions of causal responsibility, 

identification of who/what is responsible for problematic campaign 

financing or the absence of meaningful reform. 
Solution: The article mentions the solution to the problem of 

campaign financing and identifies who/what will remedy the 

problem. 
None of the above 



CODER COMMENTS/ IF ANY: 



Content Analysis of Attributes 

Code Book for Congressional Agenda 





GENERAL INFORMATION 




(filled out by researcher) 


Coded by 


Date coded 


Source: 




<> House 


<> Senate 


Text Identification # 


Week of text 



148 



SUB-ISSUE ATTRIBUTES 



Choose and circle one of the following sub-issues to represent the primary 
attribute in the text: 

<> Freedom of speech: any discussion of money as speech or 1 st 

amendment rights 
Soft Money: discussion regarding unregulated contributions from 

special interest groups (unions, corporations, etc.) 
Candidate advertising: reference to political advertisements, free 
commercial time, express advocacy issues 
Incumbency Advantage: reference to incumbent resources and 

increased access- for example: franking privileges, gerrymandering 
Contribution Limits: discussion regarding limiting campaign 
contributions & incentives to do so 
Expenditure Limits: discussion regarding limiting a candidate's 

campaign expenditures 
Political corruption: reference to American cynicism with the 

democratic process 
Legislation: discussion regarding specific campaign reform 

legislation or bills before congress- for example: 1971 FECA 
Reform proponents: reference to specific individuals such as McCain, 

Feingold, Shays, Meehan 
<> Other 

Do any of the following sub-issues represent secondary attributes in the 
text?(choose as many as apply, but DO NOT choose your primary attribute): 

<> Freedom of speech: money as speech, 1 st amendment right 
<> Soft Money: unregulated contributions from special interest 

groups (unions, corporations, etc.) 
<> Candidate advertising: advertising, express advocacy, free airtime 



149 



Incumbency Advantage: due to franking privileges, 

gerrymandering, etc. 

Contribution Limits: regarding limiting amount of contributions, 

incentives 

Expenditure Limits: regarding limiting campaign expenditures 

Political corruption: threat to democracy, pervasive American 

cynicism 

Legislation: example-1971 FECA 

Reform proponents: McCain, Feingold, Shays, Meehan 

Other 



FRAMING MECHANISMS 



(filled out by researcher) 
# of words in the speech 
Who is the speaker? 



<1> Democrat <2> Independent <3> Republican 



AFFECTIVE ATTRIBUTES 



Choose and circle one of the following to represent the tone of the text 
analyzed: 

<> Negative: generally engaging in a negative or judgmental tone, hostile 
or conflictual, using attacks or aggressive language, pessimistic or 
sarcastic evaluations of the topic and those involved (word examples: 
corrupt, disgust, frustration, chaotic) 



150 



Positive: generally making judgments in a positive or optimistic tone, 
stressing fairness and acceptance, presenting views using logical and 
non-inflammatory language, calm and sensitive demeanor (word 
examples: agree, change, compromise, helpful, benefit) 

Neutral: The absence of judgment; statements of fact; nonevaluative 
reports of related events, gatherings, or meetings. 

None of the above 



COGNITIVE ATTRIBUTES 



Do any of the following represent the cognitive structure of the speech (you 
may choose as many as apply): 



Cause: The article refers to attributions of causal responsibility, 

identification of who/what is responsible for problematic campaign 

financing or the absence of meaningful reform. 
Solution: The article mentions the solution to the problem of 

campaign financing and identifies who/what will remedy the 

problem. 
None of the above 



CODER COMMENTS/ IF ANY: 



REFERENCES 

Alexander, H.E. (1984). Financing politics (3 rd ed). Congressional 
Quarterly, Inc. 

Ball-Rokeach, S.J. & Defleur, M.L. (1976). A dependency model of mass 
media effects. Communication Research, 3 , pp. 3-21. 

Becker, L.B. (1991). Reflecting on metaphors. In J. A. Anderson (Ed.), 
Communication Yearbook, 14 , pp. 341-346. Newbury Park, CA: Sage. 

Becker, L, & McCombs, M. (1978). The role of the press in determining voter 
reactions to presidential primaries. Human Communication Research, 4 , 
pp. 301-307. 

Behr, R.L. & Iyengar, S. (1985). Television news, real-world cues, and 
change in the public agenda. Public Opinion Quarterly, 49 , pp. 38-57. 

Beniger, J. R. (1978). Media content as social indicators: The Greenfield 
index of agenda-setting. Communication Research. 5 . pp. 437-451 . 

Beniger, J.R. (1987). Toward an old new paradigm: The half century flirtation 
with mass society. Public Opinion Quarterly. 51 . pp. s47-s66. 

Benton, M., & Frazier, P.J. (1976). The agenda-setting function of the mass 
media at three levels of "information holding." Communication Research. 
3(3). pp. 261-274. 

Berelson, B. (1948). Communications and public opinion. In W. Schramm 
(Ed.), Communications in modern society, pp. 167-185. Urbana: 
University of Illinois Press. 

Berger, J. (1995, December 30). Critics of traditional media are flocking to the 
web. Editor & Publisher. 128 . p. 40 



151 



152 



Bichard, S.L. (1999, October). Exploring intermedia influences on agenda 
setting in national newspaper coverage of genetic cloning . Paper 
presented at the Florida Communication Association convention. Ocala, 
FL. 

Blumer, Herbert (1969). Collective behavior. In Alfred McClung Lee's (Ed.), 
Principles of Sociology , pp. 65-121. Harper & Row Publishers: New York. 

Bobbitt, R. (1995). An Internet primer for public relations. Public Relations 
Quarterly , Rhinebeck. 

Box, G.E.P. & Jenkins, G.M. (1976). Time series analysis: Forecasting and 
control . San Francisco: Holden-Day. 

Breed, W. (1980). Dissertations on sociology: The newspaperman, news and 
society . New York: Arno Press. 

Brosius, H. & Kepplinger, H. (1995). Killer and victim issues: Issue 
competition in the agenda-setting process of German television. 
International Journal of Public Opinion Research, 7, N3 , pp. 21 1 -231 . 

Brosius, H. & Weimann, G. (1996). Who sets the agenda?: Agenda-setting 
as a two-step flow. Communication Research. V23, N5 . pp. 561-580. 

Carter, R.F., Stamm, K.R. & Heintz-Knowles, K. (1992). Agenda-setting and 
consequentiality. Journalism Quarterly, 69 . pp. 868-877. 

Chaffee, S.H. (1972, August). Longitudinal designs for communication 
research: Cross-lagged correlations . Paper presented at the Association 
for Education in Journalism and Mass Communication, Carbondale, IL. 

Chaffee, S.H. (1973). Contingent orientations and the effects of political 
communication . Presented to the Speech Communication Association, 
New York. 

Chaffee, S.H. (1978). Presidential debates- Are they helpful to voters? 
Communication Monographs, 43 , pp. 330-346. 

Cohen, B. (1963). The press and foreign policy . Princeton: Princeton 
University Press. 

Cohen, B.C. (1965). Foreign policy in American government . Boston: Little, 
Brown. 



153 



Common Cause. (Press release, February, 18, 1988). Senate and House 
showdowns on campaign finance reform set for early '98 . 

Converse, P.E. (1972). Change in the American electorate. In A. Campbell 
and P.E. Converse (Eds.), The human meaning of social change . New 
York: Russell Sage Foundation. 

Converse, P.E. (1987). Changing conceptions of public opinion in the political 
process. Public Opinion Quarterly. 51 , pp. s12-s24. 

Cook, F.L., Tyler, T.R., Goetz, E.G., Gordon, M.T., Protess, D., Leff, DR., & 
Molotch, H.L. (1983). Media and agenda setting: Effects on the public, 
interest group leaders, policymakers and policy. Public Opinion Quarterly. 
47, pp. 16-35. 

Corbett, J. B. & Mori, M. (1999). Medicine, media, and celebrities: News 
coverage of breast cancer, 1 960-1 995. Journalism and Mass 
Communication Quarterly. 76(2 ). pp. 229-249. 

Danielian, L. & Reese, S. (1989). A closer look at intermedia influences on 
agenda setting: The cocaine issue of 1986. In Pamela J. Shoemaker's 
(Ed.), Communication Campaigns About Drugs: Government, Media, and 
the Public , pp.47-66. Hillsdale, NJ: Lawrence Erlbaum Associates. 

Davis, F.J. (1952). Crime news in Colorado newspapers. American Journal 
of Sociology. 57 , pp. 325-330. 

Davison, W. P. (1958). The public opinion process. Public Opinion 
Quarterly, 22 . pp. 91-106. 

Demers, D.P. (1996). Does personal experience in a community increase or 
decrease newspaper reading? Journalism & Mass Communication 
Quarterly. 73(2) . pp. 304-318. 

Demers, D.P., Craff, D., Choi, Y., & Pessin, B.M. (1989). Issue obtrusiveness 
and the agenda-setting effects of national network news. Communication 
Research, 16 . pp. 793-812. 

Diamond, E., Accosta, F., & Thornton, L. (1987). Is TV news hyping 
America's cocaine problem? TV Guide. February 7 , pp. 4-10. 



154 



Domke, D., Fan, DP., Fibison, M., Shah, D.V., Smith, S.S., & Watts, M.D. 
(1997). News media, candidates and issues, and public opinion in the 
1996 presidential campaign. Journalism and Mass Communication 
Quarterly, 74 , pp. 718-737. 

Donohew, L. (1983). Newswriting styles: What arouses the reader? 
Newspaper Research Journal, 3 , pp. 3-6. 

Downs, A. (1972). Up and down with ecology-The "issue-attention cycle." 
Public Interest, 28 , pp. 28-50. 

Dzwo, T. (1998, August). Crises on the cyberspace: Applying agenda-setting 
theory to on-line crisis management . Paper presented at the Association 
for Education in Journalism and Mass Communication conference. 
Baltimore, MD. 

Edelstein, A. (1993). Thinking about the criterion variable in agenda setting 
research. Journal of Communication. 43 , pp. 85-99. 

Edelstein, A., Ito, Y., & Kepplinger, H. (1989). Communication & culture: A 
comparative approach . New York: Longman. 

Edmondson, B. (1997). The Wired Bunch. American Demographics . June, 
1997. 

Einsiedel, E., Salomone, K., & Schneider, F. (1984). Crime: Effects of media 
exposure and personal experience on issue salience. Journalism 
Quarterly. 61 . pp. 131-136. 

Elliott, D. (1988). Family ties: A case study of coverage of families and friends 
during the hijacking of TWA flight 847. Political Communication and 
Persuasion. 5 . pp. 67-75. 

Entman, R. (1991). Framing: Toward a clarification of a fractured paradigm. 
Journal of Communication, 43 . pp. 51-58. 

Entman, R., & Rojecki, A. (1993). Freezing out the public: Elite and media 
framing of the U.S. anti-nuclear movement. Political Communication. 10 . 
pp. 155-173. 

Erbring, L, Goldenberg, E.N., & Miller, AH. (1980). Front page news and 
real world cues: A new look at agenda setting by the media. American 
Journal of Political Science. 24 . pp. 46-49. 



155 



Eulau, W. (1959). The role of the representative: Some empirical 

observations on the theory of Edmund Burke. American Political Science 
Review. 53 . pp. 742-756. 

Fishman, M. (1981). Crime waves as ideology. In S. Cohen & J. Young 
(Eds.), The Manufacture of News , pp.98-117. Beverly Hills, CA: Sage. 

Fiske, S., & Taylor, S. (1991). Social cognition . New York: McGraw-Hill. 

Foote, N. & Hart, C.W. (1953). Public opinion and collective behavior. 
In Muzafer Sherif & M.O. Wilson's (Eds.), Group relations at the 
crossroads . Harper & Brothers, New York. pp. 308-331 . 

Fowler, R. (1991). Language in the news: Discourse and ideology in the 
British press . New York: Routledge. 

Funkhouser, G.R. (1973). The Issues of the sixties: An exploratory study of 
the dynamics of public opinion. Public Opinion Quarterly. 37 . pp. 62-75. 

Gamson, W.A. (1988). Political discourse and collective action. In B. 

Klandermans, H. Friesi, & S. Tarrow(Eds.), International social movement 
research: A research annual. From structure to action: Comparing social 
movement research across cultures , pp. 219-244. Greenwich: JAI. 

Gamson, W.A. (1992). Talking politics . Cambridge: Cambridge University 
Press. 

Gamson, W.A., Croteau, D., Hoynes, W., & Sasson, T. (1992). Media images 
and the social construction of reality. Annual Review of Sociology. 1 8 pp 
373-393. 

Gamson, W.A., & Lasch, K.E. (1983). The political culture of social welfare 
policy. In S.E. Spiro, & E. Yuchtman-Yaar (Eds.), Evaluating the welfare 
state: Social and political perspectives , pp. 397-415. Academic Press, 
New York. 

Gamson, W.A., & Modigliani, A. (1989). Media discourse and public opinion 
on nuclear power: A constructionist approach. American Journal of 
Sociology. 95 . pp. 1-37. 

Gans, H. (1972). The politics of culture in America: A sociological analysis. In 
D. McQuail (Ed.), Sociology of Mass Communications , p. 373. 
Harmondsworth, England: Penguin Books. 



156 



Gans, H. (1979). Deciding what's news: A study of CBS Evening News, NBC 
Nightly News, Newsweek and Time . New York: Random House. 

Ghanem, S. (1996). Media coverage of crime and public opinion: An 
exploration of the second level of agenda setting . Unpublished doctoral 
dissertation. The University of Texas at Austin. 

Ghanem, S. (1997). Filling in the tapestry: The second level of agenda 
setting. In McCombs, M.E., Shaw, D.L., & Weaver, D. (Eds.), 
Communication and democracy: Exploring the intellectual frontiers in 
agenda-setting theory , pp. 3-14. Mahwah, NJ: Lawrence Erlbaum 
Associates. 

Gitlin, T. (1980). The whole world is watching: Mass media in the making and 
unmaking of the new left . Berkeley: University of California Press. 

Goffman, E. (1974). Frame analysis . Boston: Northeastern University Press. 

Gonzenbach, W.F. (1996). The media, the president, and public 

opinion: A longitudinal analysis of the drug issue, 1984-1991 . Mahwah, 
NJ: Lawrence Erlbaum Associates. 

Gonzenbach, W.F. (1992). A time-series analysis of the drug issue, 
1985-1990: The press, the president, and public opinion. International 
Journal of Public Opinion Research. 4 . pp. 126-1 47. 

Graber, D.A. (1979). Is crime news coverage excessive? Journal of 
Communication, 29 , pp. 81-92. 

Graber, D.A. (1980). Mass media and American politics . Washington: CQ 
Press. 

Groper, R. (1996, August/September). Political behavior of actors on electric 
discussion groups . Paper presented at the 92 nd Annual Meeting of 
American Political Science Association, San Francisco, CA. 

Hall, S. (1981). A world at one with itself. In S. Cohen & J. Young (Eds.), The 
manufacture of news: Social problems, deviance and the mass media , pp. 
147-156. Newbury, CA: Sage. 

Hall, S., Critcher, C, Jefferson, T., Clarke, J., & Roberts, B. (1978). Policing 
the crisis: Mugging, the state, and law and order . New York: Holmes & 
Meier. 



157 



Hendrickson, L. (1995, May). Effects of framing uniformity on the perception 
of child neglect as a "family problem. " Paper presented at the International 
Communication Association, Albuquerque, NM. 

Hester, J.B. & Gonzenbach, W.J. (1994). The environment, 1987-1991: The 
relationship of TV news, real world cues and public opinion over time . 
Paper submitted to the Communication Theory and Methodology Division, 
Association for Education in Journalism and Mass Communication 
Conference. 

Hilgartner, S., & Bosk, C.L. (1988). The rise and fall of social problems: A 
public arenas model. American Journal of Sociology, 94 , pp. 53-78. 

Hill, D.B. (1985). Viewer characteristics and agenda setting by television 
news. Public Opinion Quarterly. 49 . pp. 340-350. 

Hill, K.A. & Hughes, J.E. (1997). Computer-mediated political communication: 
The USENET and political communication. Political Communication, 
Mil. PP-3-27. 

Holsti, 0. (1969). Content analysis for the social sciences and humanities . 
Reading, MA: Addison-Wesley. 

Huckins, K. (1999). Interest-group influence on the media agenda: A case 
study. Journalism and Mass Communication Quarterly. 76(1 V pp. 76-86. 

lorio, S.H., & Huxman, S.S. (1996). Media coverage of political issues and the 
framing of personal concerns. Journal of Communication, 46(41 pp. 97- 
115. 

Iyengar, S. (1988). New directions of agenda-setting research. In James A. 
Anderson (Ed.), Communication Yearbook. 1 1 . pp. 595-602 Newbury 
Park, CA: Sage. 

Iyengar, S. (1989). How citizens think about national issues: A matter of 
responsibility. American Journal of Political Science. 33 . pp. 878-900. 

Iyengar, S. (1990). Framing responsibility for political issues: The case of 
poverty. Political Behavior. 12 . pp. 19-40. 

Iyengar, S. (1991). Is anyone responsible? How television frames political 
issues. Chicago: University of Chicago Press. 



158 



Iyengar, S., Peters, M.D., & Kinder, D.R. (1982). Experimental 

demonstrations of the 'not so minimal' consequences of television news 
programs. American Political Science Review. 76(4) . pp. 848-858. 

Iyengar, S., & Kinder, D.R. (1987). News that Matters: Television and 
American Opinion . Chicago: The University of Chicago Press. 

Iyengar, S., & Simon, A. (1993). News coverage of the gulf crisis and public 
opinion: A study of agenda-setting, priming, and framing. Communication 
Research. 20(31 pp. 365-383. 

Jacobson, G.C. (1980). Money in congressional elections . New York: Yale 
University Press. 

Jablonski, P.M. & Bichard, S.L. (1997). Unobtrusive issues and the agendas 
of the President, the press, and the public: The case of the environment, 
1987-1994. The Florida Communication Journal XXV. 2 . pp. 41-59. 

Jablonski, P.M. & Gonzenbach, W. J. (1996). Crime and agenda-setting, 
1988-1995: The relationships among the president, the press, and the 
public. Paper submitted to Journalism and Mass Communication 
Quarterly . 

Jasperson, A.E., Watts, M., Faber, R.J., & Fan, D.P. (1998). Framing the 
public agenda: Media effects on the importance of the federal budget 
deficit. Political Communication. 15 . pp. 205-224. 

Jones, T.E. (1976). The press as metropolitan monitor. Public Opinion 
Quarterly. 40 . pp. 239-244. 

Kahneman, D. (1982). The psychology of preferences. Science. 246 
pp. 136-1 42. ' 

Kahneman, D.; & Tversky, A. (1984). Choices, values, and frames. American 
Psychologist. 39 . pp.341 -350. 

Kaplan, A. (1964). The conduct of inouirv . Scranton, PA: Chandler. 

Katz, E., & Lazarsfeld, P. (1955). Personal Influence: The part plaved bv 
people in t he flow of mass communications Glencoe, III.: Free Press. 

Kinder, D.R., & Sanders, L.M. (1990). Mimicking political debate with survey 
questions: The case of white opinion on affirmative action for blacks 
Social Cognition. 8 , pp. 73-103. 



159 



Klandermans, H., & Sidney, T. (1988). International social movement 
research: A research annual. From structure to action: comparing social 
movement research across cultures . Greenwich, CT: JAI. 

Klapper, J.T. (1960). The effects of mass communication . New York: Free 
Press. 

Kosicki, G.M. (1993). Problems and opportunities in agenda-setting research. 
Journal of Communication. 43(2) . pp. 100-127. 

Krosnick, J.A., & Brannon, L.A. (1993). The impact of war on the ingredients 
of presidential evaluations: Multidimensional effects on political 
involvement. American Political Science Review. 87 . pp. 963-975. 

Krosnick, J. A., & Kinder, D.R. (1990). Altering the foundations of popular 
support for the president through priming. American Political Science 
Review. 84. pp. 497-512. 

Lambeth, E.B. (1978). Perceived influence of the press on energy policy 
making. Journalism Quarterly. 55 . pp. 11-18. 

Lang, G.E. & Lang, K. (1983). The battle for public opinion: The president, the 
press, and the polls during Watergate . New York: Columbia University 
Press. 

Lasorsa, D.L. (1997). Media agenda setting and press performance: A social 
system approach for building theory. In McCombs, M.E., Shaw, D.L, & 
Weaver, D. (Eds.), Communication and Democracy: Exploring the 
Intellectual Frontiers in Agenda-Setting Theory (pp. 155-168). Mahwah, 
NJ: Lawrence Erlbaum Associates. 

Lazarsfeld, P., Berelson, B., & Gaudet, H. (1944). The people's choice: How 
the voter m akes up his mind in a presidential campaign . New York: 
Columbia University Press. 

Left, D.R., Protess, D.L, & Brooks, S.C. (1986). Crusading journalism: 
Changing public attitudes and policy-making. Public Opinion Quarterly. 
50, pp. 300-315. 

Linsky, M. (1986). Impact: How the press affects federal policy making New 
York: W.W. Norton. 

Lippmann, W. (1922). Public Opinion . New York: Macmillan. 



160 



Lopez-Escobar, E., Llamas, J. P., McCombs, M. & Rey, F. (1998). Two levels 
of intermedia agenda setting among advertising and news agendas in the 
1995 Spanish regional elections. Political Communication. 15 . pp. 225- 
238. 

Luntz research company. (September, 1997). The Wired/Merrill Lynch Forum 
Digital Citizen Survey . The Digital Citizen, Jon Katz. 

Mackuen, MB. (1981). Social communication and the mass policy agenda. In 
M.B. Mackuen & S.L. Coombs (Eds.), More than news: Media power in 
public affairs , pp. 19-144, Newbury Park, CA: Sage. 

Mackuen, M.B. & Coombs, S.L. (1981). More than news: Media power in 
public affairs . Beverly Hills, CA: Sage. 

McCain, J. (1999, June 30). It's your country . Speech delivered in Bedford, 
New Hampshire. 

McCleary, R., & Hay, R.A. (1980). Applied times series analysis for the social 
sciences. Beverly Hills, CA: Sage. 

McCombs, M.E. (1981). The agenda-setting approach. In D.D. Nimmo & K.R. 
Sanders (Eds.), Handbook of political communication , pp. 121-140, 
Beverly Hills, CA: Sage. 

McCombs, M.E. (1992). Explorers and surveyors: Expanding strategies for 
agenda-setting research. Journalism Quarterly 69 . pp. 813-824. 

McCombs, M.E. (1995a). Capacity, diversity, and volatility of the public 
agenda: Trends from 1954-1994. Public Opinion Quarterly. 59(4 1 pp. 
495-525. 

McCombs, M.E. (1995b). The media outside and the pictures in 

our heads: Surveying the second dimension of agenda-setting . Paper 
presented at "New trends in communication," A conference of the 
University Degli Studi di Roma La Sapienza, Rome, Italy. 

McCombs, M.E. (1996, April). The pictures of politics in our heads . Keynote 
address presented at National Chengchi University, Taipei, Taiwan. 

McCombs, M.E. & Evatt, D. (1995). Issues and attributes: Exploring a new 
dimension in agenda setting. Communicacion v Sociedad 8. 1 . pp. 7-32. 



161 



McCombs, M.E. & Ghanem, S. (1998). The convergence of agenda-setting 
and framing . Unpublished manuscript. University of Texas at Austin. 

McCombs, M.E. & Gilbert, S. (1986). News influence on our pictures of the 
world. In J. Bryant & D. Zillermann (Eds.), Perspectives on media effects, 
pp. 1-15. Hillsdale, NJ: Lawrence Erlbaum Associates. 

McCombs, M.E., Llamas, J. P., Lopez-Escobar, E., & Rey, F. (1997). 
Candidate images in Spanish elections: Second-level agenda setting 
effects. Journalism and Mass Communication Quarterly, 74(4) , pp. 703- 
717. 

McCombs, M.E., Lopez-Escobar, Llamas, J. P. (2000). Setting the agenda 
of attributes in the 1996 Spanish general election. Journal of 
Communication . 

McCombs, M.E. & Shaw, D.L. (1972). The agenda setting function of mass 
media. Public Opinion Quarterly. 36 . pp. 176-187. 

McCombs, M.E., Shaw, D.L., & Weaver, D. (Eds.). (1997). Communication 
and Democracy: Exploring the Intellectual Frontiers in Agenda-Setting 
Theory. Mahwah, NJ: Lawrence Erlbaum Associates. 

McCombs, M.E. & Weaver, D. (1973). Voters' need for orientation and use of 
mass media . Presented to the International Communication Association, 
Montreal. 

McGuire, W.J. (1974). Psychological motives and communication 
gratification, In J.G. Blumler and E. Katz (Eds.) The Uses of mass 
communications: current perspectives on gratifications research . Beverly 
Hills: Sage. 

McLeod, J.M., Becker, LB., & Byrnes, J.E. (1974). Another look at the 
agenda-setting function of the press. Communication Research, 1(2) . pp. 
131-166. 

Mediamark Research Inc. (1999). MRI CvberStats. Fall 1999 . 

Miller, G. (1965). The magical number seven, plus or minus two. 
Psychological Review. 63 . pp.81 -97. 

Miller, R.E., & Wanta, W. (1996). Race as a variable in agenda setting. 
Journalism and Mass Communication Quarterly. 73(4 1 pp. 913-925. 



162 



Naples, M.J. (1979). Effective frequency: The relationship between frequency 
and advertising effectiveness . New York, NY: Association of National 
Advertisers. 

Neuman, W. R. (1990). The threshold of public attention. Public 
Opinion Quarterly. 54 . pp. 159-176. 

Nimmo, Dan D. (1978). Political communication and public opinion in 
America . Santa Monica, California: Goodyear Publishing Company, Inc. 

Noelle-Neumann, E. (1974). The Spiral of Silence. Journal of 
Communication. 24 . pp. 43-51 . 

Noelle-Neumann, E. (1983). Spieqel-Dokumentation: Personlich Keitss- 
tarke . Hamburg: Spiegel Verlag. 

Noelle-Neumann, E. (1989). Advances in the spiral of silence research. KEIO 
Communication Review. 10 . pp. 3-33. 

O'Keefe, G.J. & Reid-Nash, K. (1987). Crime news and real world blues. 
Communication Research. 14 . pp. 147-163. 

Page, B.I., & Shapiro, R.Y. (1983). Effects of public opinion on policy. 
American Political Science Review. 77 . pp. 175-1 90. 

Pan, Z., & Kosicki, G.M. (1993). Framing analysis: An approach to news 
discourse. Political Communication. 10 . pp. 59-79. 

Park, E., & Kosicki, G.M. (1995). Presidential support during the Iran-Contra 
affair: People's reasoning process and media influence. Communication 
Research. 22(2) . pp. 207-236. 

Price, V. (1992). Communication Concepts 4: public opinion . Newbury Park, 
CA: Sage. 

Price, V., & Tewksbury, D. (1995). News values and public opinion: A 
process: How political advertising and TV news prime viewers to think 
about issues and candidates. In F. Biocca (Ed), Television and political 
advertising , pp. 265-309, Hillsdale, NJ: Lawrence Erlbaum Associates. 

Pritchard, D. (1986). Homicide and bargained justice: The agenda setting 
effect of crime news on prosecutors. Public Opinion Quarterly. 50 . pp 143- 
159. 



163 



Pritchard, D., & Berkowitz, D. (1993). The limits of Agenda-setting: The 
press and political responses to crime in the United States, 1950-1980. 
International Journal of Public Opinion Research. 5(1) . pp. 86-91 . 

Protess, D.L., Cook, F.L., Curtin, T.R., Gordon, M.T., Leff, D.R., McCombs, 
M.E., & Miller, P. (1987). The impact of investigative reporting on public 
opinion and policymaking. Public Opinion Quarterly, 51 . pp. 166-185. 

Protess, D.L., Leff, D.R., Brooks, S.C., & Gordon, M.T. (1985). Uncovering 
Rape: The watchdog press and the limits of agenda-setting. Public 
Opinion Quarterly. 49 . pp. 19-37. 

Protess, D.L, & McCombs, M.E. (Eds.). (1991). Agenda-setting: Readings on 
media, public opinion and policymaking . Hillsdale, NJ: Lawrence Erlbaum 
Associates. 

Putnam, R.D. (1995). Bowling alone: America's declining social capital. 
Current. 373 . pp.3-9. 

Rafaeli, S. (1988). Interactivity: From new media to communication. In R. 
Hawkins, J. Wiemann, & S. Pingree (Eds.), Advancing communication 
science: Merging mass and interpersonal processes , pp. 110-134, 
Newbury Park, CA: Sage. 

Reese, S. & Buckalew, B. (1995). The militarism of local television: The 
routine framing of the Persian gulf war. Critical Studies in Mass 
Communication. 12 . pp. 40-59. 

Reese, S. & Danielian, L. (1989). Intermedia influence and the drug issue: 
Converging on cocaine. In Pamela J. Shoemaker's (Ed.), Communication 
campaigns about drugs: Government, media, and the public , pp.29-45, 
Hillsdale, NJ: Lawrence Erlbaum Associates. 

Roberts, M. (1992). Predicting voting behavior via the agenda-setting 
tradition. Journalism Quarterly. 69(41 pp. 878-892. 

Roberts, M. (1997). Political advertising's influence on news, the public, and 
their behavior. In McCombs, M., Shaw, D.L, & Weaver, D. (Eds.), 
Communication an d Democracy: Exploring the intellectual frontiers in 
agenda-sett ing theory , pp.85-96, Mahwah, NJ: Lawrence Erlbaum 
Associates. 



164 



Roberts, M., Wanta, W., & Dzwo, T. (1999). Agenda setting and issue 
salience online . Submitted for consideration in Journalism and Mass 
Communication Quarterly, September, 1999. 

Roberts, M. & McCombs, M. (1994). Agenda-setting and political advertising: 
Origins of the news agenda. Political Communication, 1 1 , pp.249-262. 

Rogers, E.M., & Dearing, J.W. (1987). Agenda setting research: Where has it 
been, where is it going? In James A. Anderson (Ed.), Communication 
Yearbook 1 1 . pp. 555-594. Newbury Park, CA: Sage. 

Rogers, E.M., Dearing, J.W. & Bregman, D. (1993). The anatomy of agenda- 
setting research. Journal of Communication. 43 , pp. 68-84. 

Rogers, E.M., Dearing, J.W. & Chang, S. (1991). AIDS in the 1980's: The 
agenda-setting process of a public issue. Journalism Monographs. 126 . 
pp. 1-47. 

Rucinski, D. (1992). Personalized bias in news: The potency of the particular. 
Communication Research. 19 . pp. 91-108. 

Russell, C. (1998, April). The haves and the want-nots. American 
Demographics . 

Sampat, P. (1998). Key indicators: Communication trends: Internet use grows 
exponentially. In Vital signs 1998: The environmental trends that are 
shaping our future , p. 98-99. 

Schattschneider, E.E. (1961). The semi-sovereign people . New York: Holt, 
Rinehart and Winston. 

Scheufele, D.A. (1996, August). Facets of framing: Systematic approach to an 
undertheorized area of research on mass media effects . Paper presented 
at the Association for Education in Journalism and Mass Communication 
convention, Anaheim, CA. 

Scheufele, D.A. (1999). Framing as a theory of media effects. Journal of 
Communication. 49(1) . pp. 103-122. 

Shah, D., & Domke, D. (1995, August). Manipulating the media frame of 
electoral issues: An examination of voters' issue interpretations and 
decision-makino . Paper presented at the Association for Education in 
Journalism and Mass Communication conference, Washington D.C. 



165 



Shah, D., Domke, D., & Wackman, D.B. (1996). "To thine own self be true": 
Values, framing, and voter decision making strategies. Communication 
Research, 23(5 ), pp. 509-560. 

Shah, D., Domke, D., & Wackman, D.B. (1997). Values and the vote: Linking 
issue interpretations to the process of candidate choice. Journalism and 
Mass Communication Quarterly. 74(2) , pp. 357-387. 

Shaw, D.L. & McCombs, M.E. (1989). Dealing with illicit drugs: The power 
and limits of mass media agenda setting. In Pamela J. Shoemaker's (Ed.), 
Communication campaigns about drugs: Government, media, and the 
Public , pp. 11 3-1 20, Hillsdale, NJ: Lawrence Erlbaum Associates. 

Shepard, A. (1994). The gospel of public journalism. American Journalism 
Review , pp. 28-35. 

Shoemaker, P.J. (Ed.). (1989). Communication campaigns about drugs: 
Government, media and the public . Hillsdale, NJ: Lawrence Erlbaum 
Associates. 

Shoemaker, P.J., Wanta, W., & Leggett, D. (1989). Drug coverage and public 
opinion, 1972-1986. In P.J. Shoemaker (Ed.), Communication campaigns 
about drugs: Government, media and the public , pp. 67-80, Hillsdale, NJ: 
Lawrence Erlbaum Associates. 

Sigal, L.V. (1973). Reporters and officials: The organization and politics of 
newsmaking . Lexington, MA: D.C. Heath. 

Slabach, F.G. (1998). The Constitution and campaign finance reform: An 
anthology. Durham, NC: Carolina Academic Press. 

Smith, B.A. (1997). Campaign finance reform proposals and the first 

ammendment . Testimony of Bradley A. Smith, Associate Professor of Law, 
CATO Institute, Committee on the Judiciary Subcommittee on the 
Constitution United States House of Representatives, February 27. 

Statistical Package for Social Sciences, Inc. (1988). SPSS Trends . Chicago- 
SPSS. 

Starobin, P. (1995). A generation of vipers. Columbia Journalism Review . 
pp.25-32. 



166 



Stokes, D.E. (1966). Spatial models of party competition. In A. Campbell, P. 
Converse, W. Miller, & D. Stokes (Eds.), Elections and the political order. 
pp. 161-179, New York: Wiley. 

Stone, G.C., & McCombs, M.E. (1981). Tracing the time lag in agenda- 
setting. Journalism Quarterly. 58 , pp. 151-155. 

Takeshita, T. (1997). Exploring the media's roles in defining reality: From 
issue-agenda setting to attribute-agenda setting. In McCombs, M.E., 
Shaw, D.L., & Weaver, D. (Eds), Communication and democracy: 
exploring the intellectual frontiers in agenda-setting theory, pp. 15-28, 
Mahwah, New Jersey. 

Tankard, J., Hendrickson, L, Silberman, J., Bliss, K. & Ghanem, S. 
(1991, August). Media frames: Approaches to conceptualization and 
measurement . Paper presented to the Association for Education in 
Journalism and Mass Communication, Boston, MA. 

Tversky, A., & Kahneman, D. (1981 ). The framing of decisions and the 
psychology of choice. Science. 211 . pp. 453-458. 

Tversky, A., & Kahneman, D. (1982). Judgment under uncertainty: Heuristics 
and biases. In D. Kahneman, P. Slovic, & A. Tversky (Eds.), Judgement 
under uncertainty: Heuristics and biases , pp. 3-20. Cambridge University 
Press, Cambridge. 

Walker, J. (1977). Setting the agenda in the U.S. Senate: A theory of 
problem selection. British Journal of Political Science. 7 . pp. 423-445. 

Wanta, W. & Foote, J. (1994). The President-news media relationship: A 
time series analysis of agenda-setting. Journal of Broadcasting and 
Electronic Media. 38 . pp. 437-448. 

Wanta, W., & Hu, Y. (1994). Time-lag differences in the agenda-setting 
process: An examination of five news media. International Journal of 
Public Opinion Research. 6 . pp. 225-240. 

Wanta, W., & Wu, Y. (1992). Interpersonal communication and the agenda 
setting process. Journalism Quarterly. 69(4) . pp. 847-849. 

Watt, J.H. & van den Berg, S.A. (1978). Time series analysis of alternative 
media effects theories. In B.D. Reuben (Ed.), Communication Yearbook 2 . 
pp. 215-224, New Brunswick, NJ: Transaction Books. 



167 



Watt, J.H. & van den Berg, S.A. (1981). How time dependency influences 
media effects in a community controversy. Journalism Quarterly. 58 . 
pp. 43-50. 

Weaver, D.H. (1987). Media agenda-setting and elections: Assumptions and 
implication. In David L. Paletz (Ed.), Political communication research: 
Approaches, studies, and assessments , pp. 176-1 91, Norwood, NJ: 
Abex. 

Weaver, D.H., Graber, D.A., McCombs, M.E., & Eyal, C.H. (1981). Media 
agenda-setting in a presidential election: Issues, images and interest . New 
York: Praeger. 

Weaver, D.H. & McCombs, M.E. (1978). Voters need for orientation and 
choice of candidate: Mass media and electoral decision making . Paper 
presented at the Conference of the American Association for Public 
Opinion Research, Roanoke, VA. 

Weaver, D.H., McCombs, M.E., & Spellman, C. (1975). Watergate and the 
media: A case study of agenda-setting. American Politics Quarterly. 3 
pp. 458-472. 

Weise, E. (1998, April 16). Net use doubling every 100 days. USA Today . 

Willnat, L. (1997). Agenda setting and priming: Conceptual links and 
differences. In McCombs, M.E., Shaw, D.L., & Weaver, D. (Eds.), 
Communication a nd democracy: Exploring the intellectual frontiers in 
agenda-setting theory , pp. 51-68, Mahwah, New Jersey. 

Winter, J.P., & Eyal, C.H. (1981). Agenda setting for the civil rights issue 
Public Opinion Quarterly. 45 pp. 376-383. 

Wyer, R.S., & Srull, T.K. (1986). Human cognition in its social context. 
Psychological Review. 93 . pp. 322-359. 

Wyer, R.S., & Srull, T.K. (1989). Memory and cognition in its social context 
Hillsdale, NJ: Lawrence Erlbaum Associates. 

Yaffee, R.A. & McGee, M. (2000). Introduction to time series analysis and 
forecasting. San Diego, CA: Academic Press, Inc. 

Yagade, A, & Dozier, D.M. (1990). The media agenda setting effect of 
concrete versus abstract issues. Journalism Quarterly. 67 pp. 3-10. 



168 



Yankelovich, D. (1985). How the public learns the public's business. 
Kettering Review, 9 , pp. 8-18. 

Zhu, J. (1992). Issue competition and attention distraction: A zero-sum 
theory of agenda-setting. Journalism Quarterly, V69 . pp. 825-836. 

Zhu, J.H. Watt, J.H., Snyder, LB., Yan, J., & Jiang, Y. (1993). Public issue 
priority formation: Media agenda-setting and social interaction. Journal of 
Communication, 43 , pp. 8-29. 

Zucker, H.G. (1978). The variable nature of the news media influence. In B. 
Reuben (Ed.), Communication Yearbook 2, pp. 225-240, New Brunswick, 
NJ: Transaction Books. 



BIOGRAPHICAL SKETCH 
Shannon L. Bichard (maiden name Crosby) was born in Orlando, Florida, 
on November 16, 1973. She received her bachelor's degree in organizational 
communication in 1995 from the University of Central Florida. She then went 
on to receive her master's degree in communication in 1997 also from the 
University of Central Florida. In the summer of 1997, she was married to 
Robert 0. Bichard and moved to Gainesville, Florida, to begin her doctoral 
studies at the University of Florida. After four long years of hard work, she 
will soon graduate and begin her career in the fall of 2001 as an assistant 
professor of advertising at Texas Tech University. 





















169 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope 
and quality, as a dissertation for the degree of Doctor of Philosophy. 




*t^^7 



Marilyn S. Roberts, Chair 
Associate Professor of Journalism and 
Communications 

I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope 
and quality, as a dissertation for the degree of Doctor of Philosophy. 




jc^uAjuJUj) 



John C. Sutherland 
Professor of Journalism and 
Communications 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope 
and quality, as a dissertation for the degree of Doctor of Philosophy. 




Wayne Wanta 

Associate Professor of Journalism and 
Communications 

I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope 
and quality, as a dissertation for the degree of Doctor of Philosophy. 

Leonard Tipton 
Professor of Journalism and 
Communications 

I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope 
and quality, as a dissertation for the degree of Doctor of Philosophy. 



Michael D. Martinez ^-V 
Associate Professor of \ 
Political Science 



This dissertation was submitted to the Graduate Faculty of the College of 
Journalism and Communications and to the Graduate School and was 
accepted as partial fulfillment of the requirements for the degree of Doctor of 
Philosophy. 



August, 2001 



Dean, College of Journalism and 
Communications 



Dean, Graduate School 

















LD 
1780 

POD I 










$&* 














UNIVERSITY OF FLORIDA 



3 1262 08555 3518